Datasets:

hexuan21

/

VideoFeedback_bad8k

like 0

[ "2003186_00.jpg", "2003186_01.jpg", "2003186_02.jpg", "2003186_03.jpg", "2003186_04.jpg", "2003186_05.jpg", "2003186_06.jpg", "2003186_07.jpg", "2003186_08.jpg", "2003186_09.jpg", "2003186_10.jpg", "2003186_11.jpg", "2003186_12.jpg", "2003186_13.jpg", "2003186_14.jpg", "2003186_15.jpg" ]

[ { "from": "human", "value": "\nSuppose you are an expert in judging and evaluating the quality of AI-generated videos,\nplease watch the following frames of a given video and see the text prompt for generating the video,\nthen give scores from 5 different dimensions:\n(1) visual quality: the quality of the video in terms of clearness, resolution, brightness, and color\n(2) temporal consistency, both the consistency of objects or humans and the smoothness of motion or movements\n(3) dynamic degree, the degree of dynamic changes\n(4) text-to-video alignment, the alignment between the text prompt and the video content\n(5) factual consistency, the consistency of the video content with the common-sense and factual knowledge\n\nfor each dimension, output a float number from 1.0 to 4.0,\nthe higher the number is, the better the video performs in that sub-score, \nthe lowest 1.0 means Bad, the highest 4.0 means Perfect/Real (the video is like a real video)\nHere is an output example:\nvisual quality: 3.2\ntemporal consistency: 2.7\ndynamic degree: 4.0\ntext-to-video alignment: 2.3\nfactual consistency: 1.8\n\nFor this video, the text prompt is \"car driving on a highway, dynamik lighting, night time\",\nall the frames of video are as follows:\n\n<image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> " }, { "from": "gpt", "value": "visual quality: 2\n temporal consistency: 3\n dynamic degree: 3\n text-to-video alignment: 0\n factual consistency: 2\n" } ]

[ "2002408_00.jpg", "2002408_01.jpg", "2002408_02.jpg", "2002408_03.jpg", "2002408_04.jpg", "2002408_05.jpg", "2002408_06.jpg", "2002408_07.jpg", "2002408_08.jpg", "2002408_09.jpg", "2002408_10.jpg", "2002408_11.jpg", "2002408_12.jpg", "2002408_13.jpg", "2002408_14.jpg", "2002408_15.jpg" ]

[ { "from": "human", "value": "\nSuppose you are an expert in judging and evaluating the quality of AI-generated videos,\nplease watch the following frames of a given video and see the text prompt for generating the video,\nthen give scores from 5 different dimensions:\n(1) visual quality: the quality of the video in terms of clearness, resolution, brightness, and color\n(2) temporal consistency, both the consistency of objects or humans and the smoothness of motion or movements\n(3) dynamic degree, the degree of dynamic changes\n(4) text-to-video alignment, the alignment between the text prompt and the video content\n(5) factual consistency, the consistency of the video content with the common-sense and factual knowledge\n\nfor each dimension, output a float number from 1.0 to 4.0,\nthe higher the number is, the better the video performs in that sub-score, \nthe lowest 1.0 means Bad, the highest 4.0 means Perfect/Real (the video is like a real video)\nHere is an output example:\nvisual quality: 3.2\ntemporal consistency: 2.7\ndynamic degree: 4.0\ntext-to-video alignment: 2.3\nfactual consistency: 1.8\n\nFor this video, the text prompt is \"chasing culprits through slum, speed movements,camera follows, camera zooms and out, vertigo shot, coimbatore slum,\",\nall the frames of video are as follows:\n\n<image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> " }, { "from": "gpt", "value": "visual quality: 2\n temporal consistency: 3\n dynamic degree: 3\n text-to-video alignment: 0\n factual consistency: 3\n" } ]

[ "3006076_00.jpg", "3006076_01.jpg", "3006076_02.jpg", "3006076_03.jpg", "3006076_04.jpg", "3006076_05.jpg", "3006076_06.jpg", "3006076_07.jpg", "3006076_08.jpg", "3006076_09.jpg", "3006076_10.jpg", "3006076_11.jpg", "3006076_12.jpg", "3006076_13.jpg", "3006076_14.jpg", "3006076_15.jpg" ]

[ { "from": "human", "value": "\nSuppose you are an expert in judging and evaluating the quality of AI-generated videos,\nplease watch the following frames of a given video and see the text prompt for generating the video,\nthen give scores from 5 different dimensions:\n(1) visual quality: the quality of the video in terms of clearness, resolution, brightness, and color\n(2) temporal consistency, both the consistency of objects or humans and the smoothness of motion or movements\n(3) dynamic degree, the degree of dynamic changes\n(4) text-to-video alignment, the alignment between the text prompt and the video content\n(5) factual consistency, the consistency of the video content with the common-sense and factual knowledge\n\nfor each dimension, output a float number from 1.0 to 4.0,\nthe higher the number is, the better the video performs in that sub-score, \nthe lowest 1.0 means Bad, the highest 4.0 means Perfect/Real (the video is like a real video)\nHere is an output example:\nvisual quality: 3.2\ntemporal consistency: 2.7\ndynamic degree: 4.0\ntext-to-video alignment: 2.3\nfactual consistency: 1.8\n\nFor this video, the text prompt is \"Realistic Raining from the sky, Dark clouds\",\nall the frames of video are as follows:\n\n<image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> " }, { "from": "gpt", "value": "visual quality: 1\n temporal consistency: 3\n dynamic degree: 3\n text-to-video alignment: 0\n factual consistency: 2\n" } ]

[ "4002970_00.jpg", "4002970_01.jpg", "4002970_02.jpg", "4002970_03.jpg", "4002970_04.jpg", "4002970_05.jpg", "4002970_06.jpg", "4002970_07.jpg", "4002970_08.jpg", "4002970_09.jpg", "4002970_10.jpg", "4002970_11.jpg", "4002970_12.jpg", "4002970_13.jpg", "4002970_14.jpg", "4002970_15.jpg" ]

[ { "from": "human", "value": "\nSuppose you are an expert in judging and evaluating the quality of AI-generated videos,\nplease watch the following frames of a given video and see the text prompt for generating the video,\nthen give scores from 5 different dimensions:\n(1) visual quality: the quality of the video in terms of clearness, resolution, brightness, and color\n(2) temporal consistency, both the consistency of objects or humans and the smoothness of motion or movements\n(3) dynamic degree, the degree of dynamic changes\n(4) text-to-video alignment, the alignment between the text prompt and the video content\n(5) factual consistency, the consistency of the video content with the common-sense and factual knowledge\n\nfor each dimension, output a float number from 1.0 to 4.0,\nthe higher the number is, the better the video performs in that sub-score, \nthe lowest 1.0 means Bad, the highest 4.0 means Perfect/Real (the video is like a real video)\nHere is an output example:\nvisual quality: 3.2\ntemporal consistency: 2.7\ndynamic degree: 4.0\ntext-to-video alignment: 2.3\nfactual consistency: 1.8\n\nFor this video, the text prompt is \"A Islamic Scholar sit in car\",\nall the frames of video are as follows:\n\n<image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> " }, { "from": "gpt", "value": "visual quality: 3\n temporal consistency: 2\n dynamic degree: 3\n text-to-video alignment: 0\n factual consistency: 1\n" } ]

[ "4003355_00.jpg", "4003355_01.jpg", "4003355_02.jpg", "4003355_03.jpg", "4003355_04.jpg", "4003355_05.jpg", "4003355_06.jpg", "4003355_07.jpg", "4003355_08.jpg", "4003355_09.jpg", "4003355_10.jpg", "4003355_11.jpg", "4003355_12.jpg", "4003355_13.jpg", "4003355_14.jpg", "4003355_15.jpg" ]

[ { "from": "human", "value": "\nSuppose you are an expert in judging and evaluating the quality of AI-generated videos,\nplease watch the following frames of a given video and see the text prompt for generating the video,\nthen give scores from 5 different dimensions:\n(1) visual quality: the quality of the video in terms of clearness, resolution, brightness, and color\n(2) temporal consistency, both the consistency of objects or humans and the smoothness of motion or movements\n(3) dynamic degree, the degree of dynamic changes\n(4) text-to-video alignment, the alignment between the text prompt and the video content\n(5) factual consistency, the consistency of the video content with the common-sense and factual knowledge\n\nfor each dimension, output a float number from 1.0 to 4.0,\nthe higher the number is, the better the video performs in that sub-score, \nthe lowest 1.0 means Bad, the highest 4.0 means Perfect/Real (the video is like a real video)\nHere is an output example:\nvisual quality: 3.2\ntemporal consistency: 2.7\ndynamic degree: 4.0\ntext-to-video alignment: 2.3\nfactual consistency: 1.8\n\nFor this video, the text prompt is \"the vast plain, Migrating Cattle, Sunset, Epic Amazing, Film Lens, Lens Pan, Best quality, HD\",\nall the frames of video are as follows:\n\n<image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> " }, { "from": "gpt", "value": "visual quality: 2\n temporal consistency: 2\n dynamic degree: 3\n text-to-video alignment: 0\n factual consistency: 1\n" } ]

[ "9000187_00.jpg", "9000187_01.jpg", "9000187_02.jpg", "9000187_03.jpg", "9000187_04.jpg", "9000187_05.jpg", "9000187_06.jpg", "9000187_07.jpg", "9000187_08.jpg", "9000187_09.jpg", "9000187_10.jpg", "9000187_11.jpg", "9000187_12.jpg", "9000187_13.jpg", "9000187_14.jpg", "9000187_15.jpg", "9000187_16.jpg", "9000187_17.jpg", "9000187_18.jpg", "9000187_19.jpg", "9000187_20.jpg", "9000187_21.jpg", "9000187_22.jpg", "9000187_23.jpg" ]

[ { "from": "human", "value": "\nSuppose you are an expert in judging and evaluating the quality of AI-generated videos,\nplease watch the following frames of a given video and see the text prompt for generating the video,\nthen give scores from 5 different dimensions:\n(1) visual quality: the quality of the video in terms of clearness, resolution, brightness, and color\n(2) temporal consistency, both the consistency of objects or humans and the smoothness of motion or movements\n(3) dynamic degree, the degree of dynamic changes\n(4) text-to-video alignment, the alignment between the text prompt and the video content\n(5) factual consistency, the consistency of the video content with the common-sense and factual knowledge\n\nfor each dimension, output a float number from 1.0 to 4.0,\nthe higher the number is, the better the video performs in that sub-score, \nthe lowest 1.0 means Bad, the highest 4.0 means Perfect/Real (the video is like a real video)\nHere is an output example:\nvisual quality: 3.2\ntemporal consistency: 2.7\ndynamic degree: 4.0\ntext-to-video alignment: 2.3\nfactual consistency: 1.8\n\nFor this video, the text prompt is \"Theme number 279, 3D text, appears from the middle, rotates in a circle, gold, only 279 appears, black background, golden dots flashing next to it\",\nall the frames of video are as follows:\n\n<image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> " }, { "from": "gpt", "value": "visual quality: 3\n temporal consistency: 2\n dynamic degree: 3\n text-to-video alignment: 0\n factual consistency: 1\n" } ]

[ "7003694_00.jpg", "7003694_01.jpg", "7003694_02.jpg", "7003694_03.jpg", "7003694_04.jpg", "7003694_05.jpg", "7003694_06.jpg", "7003694_07.jpg" ]

[ { "from": "human", "value": "\nSuppose you are an expert in judging and evaluating the quality of AI-generated videos,\nplease watch the following frames of a given video and see the text prompt for generating the video,\nthen give scores from 5 different dimensions:\n(1) visual quality: the quality of the video in terms of clearness, resolution, brightness, and color\n(2) temporal consistency, both the consistency of objects or humans and the smoothness of motion or movements\n(3) dynamic degree, the degree of dynamic changes\n(4) text-to-video alignment, the alignment between the text prompt and the video content\n(5) factual consistency, the consistency of the video content with the common-sense and factual knowledge\n\nfor each dimension, output a float number from 1.0 to 4.0,\nthe higher the number is, the better the video performs in that sub-score, \nthe lowest 1.0 means Bad, the highest 4.0 means Perfect/Real (the video is like a real video)\nHere is an output example:\nvisual quality: 3.2\ntemporal consistency: 2.7\ndynamic degree: 4.0\ntext-to-video alignment: 2.3\nfactual consistency: 1.8\n\nFor this video, the text prompt is \"a single flame in an illustration style, disney style, white background\",\nall the frames of video are as follows:\n\n<image> <image> <image> <image> <image> <image> <image> <image> " }, { "from": "gpt", "value": "visual quality: 2\n temporal consistency: 3\n dynamic degree: 3\n text-to-video alignment: 0\n factual consistency: 2\n" } ]

[ "3002988_00.jpg", "3002988_01.jpg", "3002988_02.jpg", "3002988_03.jpg", "3002988_04.jpg", "3002988_05.jpg", "3002988_06.jpg", "3002988_07.jpg", "3002988_08.jpg", "3002988_09.jpg", "3002988_10.jpg", "3002988_11.jpg", "3002988_12.jpg", "3002988_13.jpg", "3002988_14.jpg", "3002988_15.jpg" ]

[ { "from": "human", "value": "\nSuppose you are an expert in judging and evaluating the quality of AI-generated videos,\nplease watch the following frames of a given video and see the text prompt for generating the video,\nthen give scores from 5 different dimensions:\n(1) visual quality: the quality of the video in terms of clearness, resolution, brightness, and color\n(2) temporal consistency, both the consistency of objects or humans and the smoothness of motion or movements\n(3) dynamic degree, the degree of dynamic changes\n(4) text-to-video alignment, the alignment between the text prompt and the video content\n(5) factual consistency, the consistency of the video content with the common-sense and factual knowledge\n\nfor each dimension, output a float number from 1.0 to 4.0,\nthe higher the number is, the better the video performs in that sub-score, \nthe lowest 1.0 means Bad, the highest 4.0 means Perfect/Real (the video is like a real video)\nHere is an output example:\nvisual quality: 3.2\ntemporal consistency: 2.7\ndynamic degree: 4.0\ntext-to-video alignment: 2.3\nfactual consistency: 1.8\n\nFor this video, the text prompt is \"male rapper in Ethiopia, sitting front of his home, rapping, his girl next to him, 4k, detailed\",\nall the frames of video are as follows:\n\n<image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> " }, { "from": "gpt", "value": "visual quality: 2\n temporal consistency: 3\n dynamic degree: 2\n text-to-video alignment: 0\n factual consistency: 3\n" } ]

[ "3001971_00.jpg", "3001971_01.jpg", "3001971_02.jpg", "3001971_03.jpg", "3001971_04.jpg", "3001971_05.jpg", "3001971_06.jpg", "3001971_07.jpg", "3001971_08.jpg", "3001971_09.jpg", "3001971_10.jpg", "3001971_11.jpg", "3001971_12.jpg", "3001971_13.jpg", "3001971_14.jpg", "3001971_15.jpg" ]

[ { "from": "human", "value": "\nSuppose you are an expert in judging and evaluating the quality of AI-generated videos,\nplease watch the following frames of a given video and see the text prompt for generating the video,\nthen give scores from 5 different dimensions:\n(1) visual quality: the quality of the video in terms of clearness, resolution, brightness, and color\n(2) temporal consistency, both the consistency of objects or humans and the smoothness of motion or movements\n(3) dynamic degree, the degree of dynamic changes\n(4) text-to-video alignment, the alignment between the text prompt and the video content\n(5) factual consistency, the consistency of the video content with the common-sense and factual knowledge\n\nfor each dimension, output a float number from 1.0 to 4.0,\nthe higher the number is, the better the video performs in that sub-score, \nthe lowest 1.0 means Bad, the highest 4.0 means Perfect/Real (the video is like a real video)\nHere is an output example:\nvisual quality: 3.2\ntemporal consistency: 2.7\ndynamic degree: 4.0\ntext-to-video alignment: 2.3\nfactual consistency: 1.8\n\nFor this video, the text prompt is \"Create a video of a fictional city that is under a volcano\",\nall the frames of video are as follows:\n\n<image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> " }, { "from": "gpt", "value": "visual quality: 3\n temporal consistency: 3\n dynamic degree: 1\n text-to-video alignment: 0\n factual consistency: 3\n" } ]

[ "1003412_00.jpg", "1003412_01.jpg", "1003412_02.jpg", "1003412_03.jpg", "1003412_04.jpg", "1003412_05.jpg", "1003412_06.jpg", "1003412_07.jpg", "1003412_08.jpg", "1003412_09.jpg", "1003412_10.jpg", "1003412_11.jpg", "1003412_12.jpg", "1003412_13.jpg", "1003412_14.jpg", "1003412_15.jpg" ]

[ { "from": "human", "value": "\nSuppose you are an expert in judging and evaluating the quality of AI-generated videos,\nplease watch the following frames of a given video and see the text prompt for generating the video,\nthen give scores from 5 different dimensions:\n(1) visual quality: the quality of the video in terms of clearness, resolution, brightness, and color\n(2) temporal consistency, both the consistency of objects or humans and the smoothness of motion or movements\n(3) dynamic degree, the degree of dynamic changes\n(4) text-to-video alignment, the alignment between the text prompt and the video content\n(5) factual consistency, the consistency of the video content with the common-sense and factual knowledge\n\nfor each dimension, output a float number from 1.0 to 4.0,\nthe higher the number is, the better the video performs in that sub-score, \nthe lowest 1.0 means Bad, the highest 4.0 means Perfect/Real (the video is like a real video)\nHere is an output example:\nvisual quality: 3.2\ntemporal consistency: 2.7\ndynamic degree: 4.0\ntext-to-video alignment: 2.3\nfactual consistency: 1.8\n\nFor this video, the text prompt is \"dead bodies of villagers laying on ground due to war\",\nall the frames of video are as follows:\n\n<image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> " }, { "from": "gpt", "value": "visual quality: 1\n temporal consistency: 1\n dynamic degree: 3\n text-to-video alignment: 0\n factual consistency: 1\n" } ]

[ "1004010_00.jpg", "1004010_01.jpg", "1004010_02.jpg", "1004010_03.jpg", "1004010_04.jpg", "1004010_05.jpg", "1004010_06.jpg", "1004010_07.jpg", "1004010_08.jpg", "1004010_09.jpg", "1004010_10.jpg", "1004010_11.jpg", "1004010_12.jpg", "1004010_13.jpg", "1004010_14.jpg", "1004010_15.jpg" ]

[ { "from": "human", "value": "\nSuppose you are an expert in judging and evaluating the quality of AI-generated videos,\nplease watch the following frames of a given video and see the text prompt for generating the video,\nthen give scores from 5 different dimensions:\n(1) visual quality: the quality of the video in terms of clearness, resolution, brightness, and color\n(2) temporal consistency, both the consistency of objects or humans and the smoothness of motion or movements\n(3) dynamic degree, the degree of dynamic changes\n(4) text-to-video alignment, the alignment between the text prompt and the video content\n(5) factual consistency, the consistency of the video content with the common-sense and factual knowledge\n\nfor each dimension, output a float number from 1.0 to 4.0,\nthe higher the number is, the better the video performs in that sub-score, \nthe lowest 1.0 means Bad, the highest 4.0 means Perfect/Real (the video is like a real video)\nHere is an output example:\nvisual quality: 3.2\ntemporal consistency: 2.7\ndynamic degree: 4.0\ntext-to-video alignment: 2.3\nfactual consistency: 1.8\n\nFor this video, the text prompt is \"Spaceship Is Landing And Crashes into The Big Building And Exploding In Shanghai China At Night, GoPro Cam Footage Filter\",\nall the frames of video are as follows:\n\n<image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> " }, { "from": "gpt", "value": "visual quality: 1\n temporal consistency: 1\n dynamic degree: 3\n text-to-video alignment: 0\n factual consistency: 1\n" } ]

[ "5000096_00.jpg", "5000096_01.jpg", "5000096_02.jpg", "5000096_03.jpg", "5000096_04.jpg", "5000096_05.jpg", "5000096_06.jpg", "5000096_07.jpg", "5000096_08.jpg", "5000096_09.jpg", "5000096_10.jpg", "5000096_11.jpg", "5000096_12.jpg", "5000096_13.jpg", "5000096_14.jpg", "5000096_15.jpg" ]

[ { "from": "human", "value": "\nSuppose you are an expert in judging and evaluating the quality of AI-generated videos,\nplease watch the following frames of a given video and see the text prompt for generating the video,\nthen give scores from 5 different dimensions:\n(1) visual quality: the quality of the video in terms of clearness, resolution, brightness, and color\n(2) temporal consistency, both the consistency of objects or humans and the smoothness of motion or movements\n(3) dynamic degree, the degree of dynamic changes\n(4) text-to-video alignment, the alignment between the text prompt and the video content\n(5) factual consistency, the consistency of the video content with the common-sense and factual knowledge\n\nfor each dimension, output a float number from 1.0 to 4.0,\nthe higher the number is, the better the video performs in that sub-score, \nthe lowest 1.0 means Bad, the highest 4.0 means Perfect/Real (the video is like a real video)\nHere is an output example:\nvisual quality: 3.2\ntemporal consistency: 2.7\ndynamic degree: 4.0\ntext-to-video alignment: 2.3\nfactual consistency: 1.8\n\nFor this video, the text prompt is \"Vintage footage of a French woman eating a cheeseburger, professional film, handheld camera, close up, film grainy, gritty, desaturated, found footage 16:9\",\nall the frames of video are as follows:\n\n<image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> " }, { "from": "gpt", "value": "visual quality: 3\n temporal consistency: 2\n dynamic degree: 3\n text-to-video alignment: 0\n factual consistency: 1\n" } ]

[ "5000527_00.jpg", "5000527_01.jpg", "5000527_02.jpg", "5000527_03.jpg", "5000527_04.jpg", "5000527_05.jpg", "5000527_06.jpg", "5000527_07.jpg", "5000527_08.jpg", "5000527_09.jpg", "5000527_10.jpg", "5000527_11.jpg", "5000527_12.jpg", "5000527_13.jpg", "5000527_14.jpg", "5000527_15.jpg" ]

[ { "from": "human", "value": "\nSuppose you are an expert in judging and evaluating the quality of AI-generated videos,\nplease watch the following frames of a given video and see the text prompt for generating the video,\nthen give scores from 5 different dimensions:\n(1) visual quality: the quality of the video in terms of clearness, resolution, brightness, and color\n(2) temporal consistency, both the consistency of objects or humans and the smoothness of motion or movements\n(3) dynamic degree, the degree of dynamic changes\n(4) text-to-video alignment, the alignment between the text prompt and the video content\n(5) factual consistency, the consistency of the video content with the common-sense and factual knowledge\n\nfor each dimension, output a float number from 1.0 to 4.0,\nthe higher the number is, the better the video performs in that sub-score, \nthe lowest 1.0 means Bad, the highest 4.0 means Perfect/Real (the video is like a real video)\nHere is an output example:\nvisual quality: 3.2\ntemporal consistency: 2.7\ndynamic degree: 4.0\ntext-to-video alignment: 2.3\nfactual consistency: 1.8\n\nFor this video, the text prompt is \"3d cartoon the god speak with old man with burlap cloth\",\nall the frames of video are as follows:\n\n<image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> " }, { "from": "gpt", "value": "visual quality: 3\n temporal consistency: 2\n dynamic degree: 2\n text-to-video alignment: 0\n factual consistency: 3\n" } ]

[ "0001913_00.jpg", "0001913_01.jpg", "0001913_02.jpg", "0001913_03.jpg", "0001913_04.jpg", "0001913_05.jpg", "0001913_06.jpg", "0001913_07.jpg", "0001913_08.jpg", "0001913_09.jpg", "0001913_10.jpg", "0001913_11.jpg", "0001913_12.jpg", "0001913_13.jpg", "0001913_14.jpg", "0001913_15.jpg", "0001913_16.jpg", "0001913_17.jpg", "0001913_18.jpg", "0001913_19.jpg", "0001913_20.jpg", "0001913_21.jpg", "0001913_22.jpg", "0001913_23.jpg" ]

[ { "from": "human", "value": "\nSuppose you are an expert in judging and evaluating the quality of AI-generated videos,\nplease watch the following frames of a given video and see the text prompt for generating the video,\nthen give scores from 5 different dimensions:\n(1) visual quality: the quality of the video in terms of clearness, resolution, brightness, and color\n(2) temporal consistency, both the consistency of objects or humans and the smoothness of motion or movements\n(3) dynamic degree, the degree of dynamic changes\n(4) text-to-video alignment, the alignment between the text prompt and the video content\n(5) factual consistency, the consistency of the video content with the common-sense and factual knowledge\n\nfor each dimension, output a float number from 1.0 to 4.0,\nthe higher the number is, the better the video performs in that sub-score, \nthe lowest 1.0 means Bad, the highest 4.0 means Perfect/Real (the video is like a real video)\nHere is an output example:\nvisual quality: 3.2\ntemporal consistency: 2.7\ndynamic degree: 4.0\ntext-to-video alignment: 2.3\nfactual consistency: 1.8\n\nFor this video, the text prompt is \"a rapid sequence of arduino connected to other sensors\",\nall the frames of video are as follows:\n\n<image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> " }, { "from": "gpt", "value": "visual quality: 2\n temporal consistency: 3\n dynamic degree: 2\n text-to-video alignment: 0\n factual consistency: 3\n" } ]

[ "4001957_00.jpg", "4001957_01.jpg", "4001957_02.jpg", "4001957_03.jpg", "4001957_04.jpg", "4001957_05.jpg", "4001957_06.jpg", "4001957_07.jpg", "4001957_08.jpg", "4001957_09.jpg", "4001957_10.jpg", "4001957_11.jpg", "4001957_12.jpg", "4001957_13.jpg", "4001957_14.jpg", "4001957_15.jpg" ]

[ { "from": "human", "value": "\nSuppose you are an expert in judging and evaluating the quality of AI-generated videos,\nplease watch the following frames of a given video and see the text prompt for generating the video,\nthen give scores from 5 different dimensions:\n(1) visual quality: the quality of the video in terms of clearness, resolution, brightness, and color\n(2) temporal consistency, both the consistency of objects or humans and the smoothness of motion or movements\n(3) dynamic degree, the degree of dynamic changes\n(4) text-to-video alignment, the alignment between the text prompt and the video content\n(5) factual consistency, the consistency of the video content with the common-sense and factual knowledge\n\nfor each dimension, output a float number from 1.0 to 4.0,\nthe higher the number is, the better the video performs in that sub-score, \nthe lowest 1.0 means Bad, the highest 4.0 means Perfect/Real (the video is like a real video)\nHere is an output example:\nvisual quality: 3.2\ntemporal consistency: 2.7\ndynamic degree: 4.0\ntext-to-video alignment: 2.3\nfactual consistency: 1.8\n\nFor this video, the text prompt is \"Sharp realistic leonardo da vinci face in morphing transition to blonde model face blue eyes nose nose red lips with veins\",\nall the frames of video are as follows:\n\n<image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> " }, { "from": "gpt", "value": "visual quality: 3\n temporal consistency: 3\n dynamic degree: 2\n text-to-video alignment: 0\n factual consistency: 3\n" } ]

[ "1005126_00.jpg", "1005126_01.jpg", "1005126_02.jpg", "1005126_03.jpg", "1005126_04.jpg", "1005126_05.jpg", "1005126_06.jpg", "1005126_07.jpg", "1005126_08.jpg", "1005126_09.jpg", "1005126_10.jpg", "1005126_11.jpg", "1005126_12.jpg", "1005126_13.jpg", "1005126_14.jpg", "1005126_15.jpg" ]

[ { "from": "human", "value": "\nSuppose you are an expert in judging and evaluating the quality of AI-generated videos,\nplease watch the following frames of a given video and see the text prompt for generating the video,\nthen give scores from 5 different dimensions:\n(1) visual quality: the quality of the video in terms of clearness, resolution, brightness, and color\n(2) temporal consistency, both the consistency of objects or humans and the smoothness of motion or movements\n(3) dynamic degree, the degree of dynamic changes\n(4) text-to-video alignment, the alignment between the text prompt and the video content\n(5) factual consistency, the consistency of the video content with the common-sense and factual knowledge\n\nfor each dimension, output a float number from 1.0 to 4.0,\nthe higher the number is, the better the video performs in that sub-score, \nthe lowest 1.0 means Bad, the highest 4.0 means Perfect/Real (the video is like a real video)\nHere is an output example:\nvisual quality: 3.2\ntemporal consistency: 2.7\ndynamic degree: 4.0\ntext-to-video alignment: 2.3\nfactual consistency: 1.8\n\nFor this video, the text prompt is \"a tour in the Ritz Paris hotel with beautiful hall, rooms and breakfast\",\nall the frames of video are as follows:\n\n<image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> " }, { "from": "gpt", "value": "visual quality: 1\n temporal consistency: 1\n dynamic degree: 3\n text-to-video alignment: 0\n factual consistency: 1\n" } ]

[ "1002890_00.jpg", "1002890_01.jpg", "1002890_02.jpg", "1002890_03.jpg", "1002890_04.jpg", "1002890_05.jpg", "1002890_06.jpg", "1002890_07.jpg", "1002890_08.jpg", "1002890_09.jpg", "1002890_10.jpg", "1002890_11.jpg", "1002890_12.jpg", "1002890_13.jpg", "1002890_14.jpg", "1002890_15.jpg" ]

[ { "from": "human", "value": "\nSuppose you are an expert in judging and evaluating the quality of AI-generated videos,\nplease watch the following frames of a given video and see the text prompt for generating the video,\nthen give scores from 5 different dimensions:\n(1) visual quality: the quality of the video in terms of clearness, resolution, brightness, and color\n(2) temporal consistency, both the consistency of objects or humans and the smoothness of motion or movements\n(3) dynamic degree, the degree of dynamic changes\n(4) text-to-video alignment, the alignment between the text prompt and the video content\n(5) factual consistency, the consistency of the video content with the common-sense and factual knowledge\n\nfor each dimension, output a float number from 1.0 to 4.0,\nthe higher the number is, the better the video performs in that sub-score, \nthe lowest 1.0 means Bad, the highest 4.0 means Perfect/Real (the video is like a real video)\nHere is an output example:\nvisual quality: 3.2\ntemporal consistency: 2.7\ndynamic degree: 4.0\ntext-to-video alignment: 2.3\nfactual consistency: 1.8\n\nFor this video, the text prompt is \"Liminal space hallway creepy unsettling strange surrealism dreamcore Message: THE (Font: RETRO)\",\nall the frames of video are as follows:\n\n<image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> " }, { "from": "gpt", "value": "visual quality: 1\n temporal consistency: 2\n dynamic degree: 3\n text-to-video alignment: 0\n factual consistency: 1\n" } ]

[ "8002552_00.jpg", "8002552_01.jpg", "8002552_02.jpg", "8002552_03.jpg", "8002552_04.jpg", "8002552_05.jpg", "8002552_06.jpg", "8002552_07.jpg", "8002552_08.jpg", "8002552_09.jpg", "8002552_10.jpg", "8002552_11.jpg", "8002552_12.jpg", "8002552_13.jpg", "8002552_14.jpg", "8002552_15.jpg", "8002552_16.jpg", "8002552_17.jpg", "8002552_18.jpg", "8002552_19.jpg", "8002552_20.jpg", "8002552_21.jpg", "8002552_22.jpg", "8002552_23.jpg" ]

[ { "from": "human", "value": "\nSuppose you are an expert in judging and evaluating the quality of AI-generated videos,\nplease watch the following frames of a given video and see the text prompt for generating the video,\nthen give scores from 5 different dimensions:\n(1) visual quality: the quality of the video in terms of clearness, resolution, brightness, and color\n(2) temporal consistency, both the consistency of objects or humans and the smoothness of motion or movements\n(3) dynamic degree, the degree of dynamic changes\n(4) text-to-video alignment, the alignment between the text prompt and the video content\n(5) factual consistency, the consistency of the video content with the common-sense and factual knowledge\n\nfor each dimension, output a float number from 1.0 to 4.0,\nthe higher the number is, the better the video performs in that sub-score, \nthe lowest 1.0 means Bad, the highest 4.0 means Perfect/Real (the video is like a real video)\nHere is an output example:\nvisual quality: 3.2\ntemporal consistency: 2.7\ndynamic degree: 4.0\ntext-to-video alignment: 2.3\nfactual consistency: 1.8\n\nFor this video, the text prompt is \"citizens of all ages are desperate, in a street of a big City by night\",\nall the frames of video are as follows:\n\n<image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> " }, { "from": "gpt", "value": "visual quality: 1\n temporal consistency: 2\n dynamic degree: 3\n text-to-video alignment: 0\n factual consistency: 1\n" } ]

[ "5000397_00.jpg", "5000397_01.jpg", "5000397_02.jpg", "5000397_03.jpg", "5000397_04.jpg", "5000397_05.jpg", "5000397_06.jpg", "5000397_07.jpg", "5000397_08.jpg", "5000397_09.jpg", "5000397_10.jpg", "5000397_11.jpg", "5000397_12.jpg", "5000397_13.jpg", "5000397_14.jpg", "5000397_15.jpg" ]

[ { "from": "human", "value": "\nSuppose you are an expert in judging and evaluating the quality of AI-generated videos,\nplease watch the following frames of a given video and see the text prompt for generating the video,\nthen give scores from 5 different dimensions:\n(1) visual quality: the quality of the video in terms of clearness, resolution, brightness, and color\n(2) temporal consistency, both the consistency of objects or humans and the smoothness of motion or movements\n(3) dynamic degree, the degree of dynamic changes\n(4) text-to-video alignment, the alignment between the text prompt and the video content\n(5) factual consistency, the consistency of the video content with the common-sense and factual knowledge\n\nfor each dimension, output a float number from 1.0 to 4.0,\nthe higher the number is, the better the video performs in that sub-score, \nthe lowest 1.0 means Bad, the highest 4.0 means Perfect/Real (the video is like a real video)\nHere is an output example:\nvisual quality: 3.2\ntemporal consistency: 2.7\ndynamic degree: 4.0\ntext-to-video alignment: 2.3\nfactual consistency: 1.8\n\nFor this video, the text prompt is \"six lovely cat walking in the tree\",\nall the frames of video are as follows:\n\n<image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> " }, { "from": "gpt", "value": "visual quality: 3\n temporal consistency: 1\n dynamic degree: 3\n text-to-video alignment: 0\n factual consistency: 1\n" } ]

[ "0001588_00.jpg", "0001588_01.jpg", "0001588_02.jpg", "0001588_03.jpg", "0001588_04.jpg", "0001588_05.jpg", "0001588_06.jpg", "0001588_07.jpg", "0001588_08.jpg", "0001588_09.jpg", "0001588_10.jpg", "0001588_11.jpg", "0001588_12.jpg", "0001588_13.jpg", "0001588_14.jpg", "0001588_15.jpg", "0001588_16.jpg", "0001588_17.jpg", "0001588_18.jpg", "0001588_19.jpg", "0001588_20.jpg", "0001588_21.jpg", "0001588_22.jpg", "0001588_23.jpg" ]

[ { "from": "human", "value": "\nSuppose you are an expert in judging and evaluating the quality of AI-generated videos,\nplease watch the following frames of a given video and see the text prompt for generating the video,\nthen give scores from 5 different dimensions:\n(1) visual quality: the quality of the video in terms of clearness, resolution, brightness, and color\n(2) temporal consistency, both the consistency of objects or humans and the smoothness of motion or movements\n(3) dynamic degree, the degree of dynamic changes\n(4) text-to-video alignment, the alignment between the text prompt and the video content\n(5) factual consistency, the consistency of the video content with the common-sense and factual knowledge\n\nfor each dimension, output a float number from 1.0 to 4.0,\nthe higher the number is, the better the video performs in that sub-score, \nthe lowest 1.0 means Bad, the highest 4.0 means Perfect/Real (the video is like a real video)\nHere is an output example:\nvisual quality: 3.2\ntemporal consistency: 2.7\ndynamic degree: 4.0\ntext-to-video alignment: 2.3\nfactual consistency: 1.8\n\nFor this video, the text prompt is \":very seductive robot women wearing a bikinnis walking together\",\nall the frames of video are as follows:\n\n<image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> " }, { "from": "gpt", "value": "visual quality: 2\n temporal consistency: 3\n dynamic degree: 2\n text-to-video alignment: 0\n factual consistency: 3\n" } ]

[ "2002199_00.jpg", "2002199_01.jpg", "2002199_02.jpg", "2002199_03.jpg", "2002199_04.jpg", "2002199_05.jpg", "2002199_06.jpg", "2002199_07.jpg", "2002199_08.jpg", "2002199_09.jpg", "2002199_10.jpg", "2002199_11.jpg", "2002199_12.jpg", "2002199_13.jpg", "2002199_14.jpg", "2002199_15.jpg" ]

[ { "from": "human", "value": "\nSuppose you are an expert in judging and evaluating the quality of AI-generated videos,\nplease watch the following frames of a given video and see the text prompt for generating the video,\nthen give scores from 5 different dimensions:\n(1) visual quality: the quality of the video in terms of clearness, resolution, brightness, and color\n(2) temporal consistency, both the consistency of objects or humans and the smoothness of motion or movements\n(3) dynamic degree, the degree of dynamic changes\n(4) text-to-video alignment, the alignment between the text prompt and the video content\n(5) factual consistency, the consistency of the video content with the common-sense and factual knowledge\n\nfor each dimension, output a float number from 1.0 to 4.0,\nthe higher the number is, the better the video performs in that sub-score, \nthe lowest 1.0 means Bad, the highest 4.0 means Perfect/Real (the video is like a real video)\nHere is an output example:\nvisual quality: 3.2\ntemporal consistency: 2.7\ndynamic degree: 4.0\ntext-to-video alignment: 2.3\nfactual consistency: 1.8\n\nFor this video, the text prompt is \"cracken is coming, view from the ship\",\nall the frames of video are as follows:\n\n<image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> " }, { "from": "gpt", "value": "visual quality: 3\n temporal consistency: 3\n dynamic degree: 3\n text-to-video alignment: 0\n factual consistency: 1\n" } ]

[ "1005228_00.jpg", "1005228_01.jpg", "1005228_02.jpg", "1005228_03.jpg", "1005228_04.jpg", "1005228_05.jpg", "1005228_06.jpg", "1005228_07.jpg", "1005228_08.jpg", "1005228_09.jpg", "1005228_10.jpg", "1005228_11.jpg", "1005228_12.jpg", "1005228_13.jpg", "1005228_14.jpg", "1005228_15.jpg" ]

[ { "from": "human", "value": "\nSuppose you are an expert in judging and evaluating the quality of AI-generated videos,\nplease watch the following frames of a given video and see the text prompt for generating the video,\nthen give scores from 5 different dimensions:\n(1) visual quality: the quality of the video in terms of clearness, resolution, brightness, and color\n(2) temporal consistency, both the consistency of objects or humans and the smoothness of motion or movements\n(3) dynamic degree, the degree of dynamic changes\n(4) text-to-video alignment, the alignment between the text prompt and the video content\n(5) factual consistency, the consistency of the video content with the common-sense and factual knowledge\n\nfor each dimension, output a float number from 1.0 to 4.0,\nthe higher the number is, the better the video performs in that sub-score, \nthe lowest 1.0 means Bad, the highest 4.0 means Perfect/Real (the video is like a real video)\nHere is an output example:\nvisual quality: 3.2\ntemporal consistency: 2.7\ndynamic degree: 4.0\ntext-to-video alignment: 2.3\nfactual consistency: 1.8\n\nFor this video, the text prompt is \"a long hair guy who wear shorts and have lean body who is riding on a Jet ski on sea and suddenly he jump from that jet ski and fly on sea\",\nall the frames of video are as follows:\n\n<image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> " }, { "from": "gpt", "value": "visual quality: 2\n temporal consistency: 1\n dynamic degree: 3\n text-to-video alignment: 0\n factual consistency: 1\n" } ]

[ "4003199_00.jpg", "4003199_01.jpg", "4003199_02.jpg", "4003199_03.jpg", "4003199_04.jpg", "4003199_05.jpg", "4003199_06.jpg", "4003199_07.jpg", "4003199_08.jpg", "4003199_09.jpg", "4003199_10.jpg", "4003199_11.jpg", "4003199_12.jpg", "4003199_13.jpg", "4003199_14.jpg", "4003199_15.jpg" ]

[ { "from": "human", "value": "\nSuppose you are an expert in judging and evaluating the quality of AI-generated videos,\nplease watch the following frames of a given video and see the text prompt for generating the video,\nthen give scores from 5 different dimensions:\n(1) visual quality: the quality of the video in terms of clearness, resolution, brightness, and color\n(2) temporal consistency, both the consistency of objects or humans and the smoothness of motion or movements\n(3) dynamic degree, the degree of dynamic changes\n(4) text-to-video alignment, the alignment between the text prompt and the video content\n(5) factual consistency, the consistency of the video content with the common-sense and factual knowledge\n\nfor each dimension, output a float number from 1.0 to 4.0,\nthe higher the number is, the better the video performs in that sub-score, \nthe lowest 1.0 means Bad, the highest 4.0 means Perfect/Real (the video is like a real video)\nHere is an output example:\nvisual quality: 3.2\ntemporal consistency: 2.7\ndynamic degree: 4.0\ntext-to-video alignment: 2.3\nfactual consistency: 1.8\n\nFor this video, the text prompt is \"A man who looks like Tom Hanks lifts the girl who looks like Anne Hathaway is shouting.\",\nall the frames of video are as follows:\n\n<image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> " }, { "from": "gpt", "value": "visual quality: 1\n temporal consistency: 3\n dynamic degree: 1\n text-to-video alignment: 0\n factual consistency: 3\n" } ]

[ "8000038_00.jpg", "8000038_01.jpg", "8000038_02.jpg", "8000038_03.jpg", "8000038_04.jpg", "8000038_05.jpg", "8000038_06.jpg", "8000038_07.jpg", "8000038_08.jpg", "8000038_09.jpg", "8000038_10.jpg", "8000038_11.jpg", "8000038_12.jpg", "8000038_13.jpg", "8000038_14.jpg", "8000038_15.jpg", "8000038_16.jpg", "8000038_17.jpg", "8000038_18.jpg", "8000038_19.jpg", "8000038_20.jpg", "8000038_21.jpg", "8000038_22.jpg", "8000038_23.jpg" ]

[ { "from": "human", "value": "\nSuppose you are an expert in judging and evaluating the quality of AI-generated videos,\nplease watch the following frames of a given video and see the text prompt for generating the video,\nthen give scores from 5 different dimensions:\n(1) visual quality: the quality of the video in terms of clearness, resolution, brightness, and color\n(2) temporal consistency, both the consistency of objects or humans and the smoothness of motion or movements\n(3) dynamic degree, the degree of dynamic changes\n(4) text-to-video alignment, the alignment between the text prompt and the video content\n(5) factual consistency, the consistency of the video content with the common-sense and factual knowledge\n\nfor each dimension, output a float number from 1.0 to 4.0,\nthe higher the number is, the better the video performs in that sub-score, \nthe lowest 1.0 means Bad, the highest 4.0 means Perfect/Real (the video is like a real video)\nHere is an output example:\nvisual quality: 3.2\ntemporal consistency: 2.7\ndynamic degree: 4.0\ntext-to-video alignment: 2.3\nfactual consistency: 1.8\n\nFor this video, the text prompt is \"Dust particles drift through the light of the stained glass window\",\nall the frames of video are as follows:\n\n<image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> " }, { "from": "gpt", "value": "visual quality: 1\n temporal consistency: 1\n dynamic degree: 3\n text-to-video alignment: 0\n factual consistency: 1\n" } ]

[ "7000932_00.jpg", "7000932_01.jpg", "7000932_02.jpg", "7000932_03.jpg", "7000932_04.jpg", "7000932_05.jpg", "7000932_06.jpg", "7000932_07.jpg" ]

[ { "from": "human", "value": "\nSuppose you are an expert in judging and evaluating the quality of AI-generated videos,\nplease watch the following frames of a given video and see the text prompt for generating the video,\nthen give scores from 5 different dimensions:\n(1) visual quality: the quality of the video in terms of clearness, resolution, brightness, and color\n(2) temporal consistency, both the consistency of objects or humans and the smoothness of motion or movements\n(3) dynamic degree, the degree of dynamic changes\n(4) text-to-video alignment, the alignment between the text prompt and the video content\n(5) factual consistency, the consistency of the video content with the common-sense and factual knowledge\n\nfor each dimension, output a float number from 1.0 to 4.0,\nthe higher the number is, the better the video performs in that sub-score, \nthe lowest 1.0 means Bad, the highest 4.0 means Perfect/Real (the video is like a real video)\nHere is an output example:\nvisual quality: 3.2\ntemporal consistency: 2.7\ndynamic degree: 4.0\ntext-to-video alignment: 2.3\nfactual consistency: 1.8\n\nFor this video, the text prompt is \"havy rainy night, the tent is burning small firewood and there\",\nall the frames of video are as follows:\n\n<image> <image> <image> <image> <image> <image> <image> <image> " }, { "from": "gpt", "value": "visual quality: 2\n temporal consistency: 2\n dynamic degree: 3\n text-to-video alignment: 0\n factual consistency: 2\n" } ]

[ "3003222_00.jpg", "3003222_01.jpg", "3003222_02.jpg", "3003222_03.jpg", "3003222_04.jpg", "3003222_05.jpg", "3003222_06.jpg", "3003222_07.jpg", "3003222_08.jpg", "3003222_09.jpg", "3003222_10.jpg", "3003222_11.jpg", "3003222_12.jpg", "3003222_13.jpg", "3003222_14.jpg", "3003222_15.jpg" ]

[ { "from": "human", "value": "\nSuppose you are an expert in judging and evaluating the quality of AI-generated videos,\nplease watch the following frames of a given video and see the text prompt for generating the video,\nthen give scores from 5 different dimensions:\n(1) visual quality: the quality of the video in terms of clearness, resolution, brightness, and color\n(2) temporal consistency, both the consistency of objects or humans and the smoothness of motion or movements\n(3) dynamic degree, the degree of dynamic changes\n(4) text-to-video alignment, the alignment between the text prompt and the video content\n(5) factual consistency, the consistency of the video content with the common-sense and factual knowledge\n\nfor each dimension, output a float number from 1.0 to 4.0,\nthe higher the number is, the better the video performs in that sub-score, \nthe lowest 1.0 means Bad, the highest 4.0 means Perfect/Real (the video is like a real video)\nHere is an output example:\nvisual quality: 3.2\ntemporal consistency: 2.7\ndynamic degree: 4.0\ntext-to-video alignment: 2.3\nfactual consistency: 1.8\n\nFor this video, the text prompt is \"Han solo and princess leia running from storm troopers\",\nall the frames of video are as follows:\n\n<image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> " }, { "from": "gpt", "value": "visual quality: 1\n temporal consistency: 2\n dynamic degree: 3\n text-to-video alignment: 0\n factual consistency: 1\n" } ]

[ "0003390_00.jpg", "0003390_01.jpg", "0003390_02.jpg", "0003390_03.jpg", "0003390_04.jpg", "0003390_05.jpg", "0003390_06.jpg", "0003390_07.jpg", "0003390_08.jpg", "0003390_09.jpg", "0003390_10.jpg", "0003390_11.jpg", "0003390_12.jpg", "0003390_13.jpg", "0003390_14.jpg", "0003390_15.jpg", "0003390_16.jpg", "0003390_17.jpg", "0003390_18.jpg", "0003390_19.jpg", "0003390_20.jpg", "0003390_21.jpg", "0003390_22.jpg", "0003390_23.jpg" ]

[ { "from": "human", "value": "\nSuppose you are an expert in judging and evaluating the quality of AI-generated videos,\nplease watch the following frames of a given video and see the text prompt for generating the video,\nthen give scores from 5 different dimensions:\n(1) visual quality: the quality of the video in terms of clearness, resolution, brightness, and color\n(2) temporal consistency, both the consistency of objects or humans and the smoothness of motion or movements\n(3) dynamic degree, the degree of dynamic changes\n(4) text-to-video alignment, the alignment between the text prompt and the video content\n(5) factual consistency, the consistency of the video content with the common-sense and factual knowledge\n\nfor each dimension, output a float number from 1.0 to 4.0,\nthe higher the number is, the better the video performs in that sub-score, \nthe lowest 1.0 means Bad, the highest 4.0 means Perfect/Real (the video is like a real video)\nHere is an output example:\nvisual quality: 3.2\ntemporal consistency: 2.7\ndynamic degree: 4.0\ntext-to-video alignment: 2.3\nfactual consistency: 1.8\n\nFor this video, the text prompt is \"a pizza on the bottom of a donut, front view\",\nall the frames of video are as follows:\n\n<image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> " }, { "from": "gpt", "value": "visual quality: 3\n temporal consistency: 3\n dynamic degree: 3\n text-to-video alignment: 0\n factual consistency: 1\n" } ]

[ "6001571_00.jpg", "6001571_01.jpg", "6001571_02.jpg", "6001571_03.jpg", "6001571_04.jpg", "6001571_05.jpg", "6001571_06.jpg", "6001571_07.jpg", "6001571_08.jpg", "6001571_09.jpg", "6001571_10.jpg", "6001571_11.jpg", "6001571_12.jpg", "6001571_13.jpg", "6001571_14.jpg", "6001571_15.jpg" ]

[ { "from": "human", "value": "\nSuppose you are an expert in judging and evaluating the quality of AI-generated videos,\nplease watch the following frames of a given video and see the text prompt for generating the video,\nthen give scores from 5 different dimensions:\n(1) visual quality: the quality of the video in terms of clearness, resolution, brightness, and color\n(2) temporal consistency, both the consistency of objects or humans and the smoothness of motion or movements\n(3) dynamic degree, the degree of dynamic changes\n(4) text-to-video alignment, the alignment between the text prompt and the video content\n(5) factual consistency, the consistency of the video content with the common-sense and factual knowledge\n\nfor each dimension, output a float number from 1.0 to 4.0,\nthe higher the number is, the better the video performs in that sub-score, \nthe lowest 1.0 means Bad, the highest 4.0 means Perfect/Real (the video is like a real video)\nHere is an output example:\nvisual quality: 3.2\ntemporal consistency: 2.7\ndynamic degree: 4.0\ntext-to-video alignment: 2.3\nfactual consistency: 1.8\n\nFor this video, the text prompt is \"Imaginary portrayal of Einstein surrounded by equations and thought bubbles with sketches of his ideas.\",\nall the frames of video are as follows:\n\n<image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> " }, { "from": "gpt", "value": "visual quality: 1\n temporal consistency: 2\n dynamic degree: 3\n text-to-video alignment: 0\n factual consistency: 1\n" } ]

[ "6001631_00.jpg", "6001631_01.jpg", "6001631_02.jpg", "6001631_03.jpg", "6001631_04.jpg", "6001631_05.jpg", "6001631_06.jpg", "6001631_07.jpg", "6001631_08.jpg", "6001631_09.jpg", "6001631_10.jpg", "6001631_11.jpg", "6001631_12.jpg", "6001631_13.jpg", "6001631_14.jpg", "6001631_15.jpg" ]

[ { "from": "human", "value": "\nSuppose you are an expert in judging and evaluating the quality of AI-generated videos,\nplease watch the following frames of a given video and see the text prompt for generating the video,\nthen give scores from 5 different dimensions:\n(1) visual quality: the quality of the video in terms of clearness, resolution, brightness, and color\n(2) temporal consistency, both the consistency of objects or humans and the smoothness of motion or movements\n(3) dynamic degree, the degree of dynamic changes\n(4) text-to-video alignment, the alignment between the text prompt and the video content\n(5) factual consistency, the consistency of the video content with the common-sense and factual knowledge\n\nfor each dimension, output a float number from 1.0 to 4.0,\nthe higher the number is, the better the video performs in that sub-score, \nthe lowest 1.0 means Bad, the highest 4.0 means Perfect/Real (the video is like a real video)\nHere is an output example:\nvisual quality: 3.2\ntemporal consistency: 2.7\ndynamic degree: 4.0\ntext-to-video alignment: 2.3\nfactual consistency: 1.8\n\nFor this video, the text prompt is \"a dog jumping in the air chasing butterflies on a sandy beach\",\nall the frames of video are as follows:\n\n<image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> " }, { "from": "gpt", "value": "visual quality: 1\n temporal consistency: 2\n dynamic degree: 3\n text-to-video alignment: 0\n factual consistency: 1\n" } ]

[ "3003769_00.jpg", "3003769_01.jpg", "3003769_02.jpg", "3003769_03.jpg", "3003769_04.jpg", "3003769_05.jpg", "3003769_06.jpg", "3003769_07.jpg", "3003769_08.jpg", "3003769_09.jpg", "3003769_10.jpg", "3003769_11.jpg", "3003769_12.jpg", "3003769_13.jpg", "3003769_14.jpg", "3003769_15.jpg" ]

[ { "from": "human", "value": "\nSuppose you are an expert in judging and evaluating the quality of AI-generated videos,\nplease watch the following frames of a given video and see the text prompt for generating the video,\nthen give scores from 5 different dimensions:\n(1) visual quality: the quality of the video in terms of clearness, resolution, brightness, and color\n(2) temporal consistency, both the consistency of objects or humans and the smoothness of motion or movements\n(3) dynamic degree, the degree of dynamic changes\n(4) text-to-video alignment, the alignment between the text prompt and the video content\n(5) factual consistency, the consistency of the video content with the common-sense and factual knowledge\n\nfor each dimension, output a float number from 1.0 to 4.0,\nthe higher the number is, the better the video performs in that sub-score, \nthe lowest 1.0 means Bad, the highest 4.0 means Perfect/Real (the video is like a real video)\nHere is an output example:\nvisual quality: 3.2\ntemporal consistency: 2.7\ndynamic degree: 4.0\ntext-to-video alignment: 2.3\nfactual consistency: 1.8\n\nFor this video, the text prompt is \"epic ufo in blue theme fling in stormy sky and beeming blue light on land green screen\",\nall the frames of video are as follows:\n\n<image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> " }, { "from": "gpt", "value": "visual quality: 1\n temporal consistency: 2\n dynamic degree: 1\n text-to-video alignment: 0\n factual consistency: 2\n" } ]

[ "0002198_00.jpg", "0002198_01.jpg", "0002198_02.jpg", "0002198_03.jpg", "0002198_04.jpg", "0002198_05.jpg", "0002198_06.jpg", "0002198_07.jpg", "0002198_08.jpg", "0002198_09.jpg", "0002198_10.jpg", "0002198_11.jpg", "0002198_12.jpg", "0002198_13.jpg", "0002198_14.jpg", "0002198_15.jpg", "0002198_16.jpg", "0002198_17.jpg", "0002198_18.jpg", "0002198_19.jpg", "0002198_20.jpg", "0002198_21.jpg", "0002198_22.jpg", "0002198_23.jpg" ]

[ { "from": "human", "value": "\nSuppose you are an expert in judging and evaluating the quality of AI-generated videos,\nplease watch the following frames of a given video and see the text prompt for generating the video,\nthen give scores from 5 different dimensions:\n(1) visual quality: the quality of the video in terms of clearness, resolution, brightness, and color\n(2) temporal consistency, both the consistency of objects or humans and the smoothness of motion or movements\n(3) dynamic degree, the degree of dynamic changes\n(4) text-to-video alignment, the alignment between the text prompt and the video content\n(5) factual consistency, the consistency of the video content with the common-sense and factual knowledge\n\nfor each dimension, output a float number from 1.0 to 4.0,\nthe higher the number is, the better the video performs in that sub-score, \nthe lowest 1.0 means Bad, the highest 4.0 means Perfect/Real (the video is like a real video)\nHere is an output example:\nvisual quality: 3.2\ntemporal consistency: 2.7\ndynamic degree: 4.0\ntext-to-video alignment: 2.3\nfactual consistency: 1.8\n\nFor this video, the text prompt is \"The camera rotates around a large stack of vintage televisions all showing different programs — 1950s sci-fi movies, horror movies, news, static, a 1970s sitcom, etc, set inside a large New York museum gallery.\",\nall the frames of video are as follows:\n\n<image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> " }, { "from": "gpt", "value": "visual quality: 3\n temporal consistency: 2\n dynamic degree: 3\n text-to-video alignment: 0\n factual consistency: 3\n" } ]

[ "3005850_00.jpg", "3005850_01.jpg", "3005850_02.jpg", "3005850_03.jpg", "3005850_04.jpg", "3005850_05.jpg", "3005850_06.jpg", "3005850_07.jpg", "3005850_08.jpg", "3005850_09.jpg", "3005850_10.jpg", "3005850_11.jpg", "3005850_12.jpg", "3005850_13.jpg", "3005850_14.jpg", "3005850_15.jpg" ]

[ { "from": "human", "value": "\nSuppose you are an expert in judging and evaluating the quality of AI-generated videos,\nplease watch the following frames of a given video and see the text prompt for generating the video,\nthen give scores from 5 different dimensions:\n(1) visual quality: the quality of the video in terms of clearness, resolution, brightness, and color\n(2) temporal consistency, both the consistency of objects or humans and the smoothness of motion or movements\n(3) dynamic degree, the degree of dynamic changes\n(4) text-to-video alignment, the alignment between the text prompt and the video content\n(5) factual consistency, the consistency of the video content with the common-sense and factual knowledge\n\nfor each dimension, output a float number from 1.0 to 4.0,\nthe higher the number is, the better the video performs in that sub-score, \nthe lowest 1.0 means Bad, the highest 4.0 means Perfect/Real (the video is like a real video)\nHere is an output example:\nvisual quality: 3.2\ntemporal consistency: 2.7\ndynamic degree: 4.0\ntext-to-video alignment: 2.3\nfactual consistency: 1.8\n\nFor this video, the text prompt is \"image:64656113.png, timelapse girl dancing, background fire Animations\",\nall the frames of video are as follows:\n\n<image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> " }, { "from": "gpt", "value": "visual quality: 3\n temporal consistency: 3\n dynamic degree: 1\n text-to-video alignment: 0\n factual consistency: 3\n" } ]

[ "4003951_00.jpg", "4003951_01.jpg", "4003951_02.jpg", "4003951_03.jpg", "4003951_04.jpg", "4003951_05.jpg", "4003951_06.jpg", "4003951_07.jpg", "4003951_08.jpg", "4003951_09.jpg", "4003951_10.jpg", "4003951_11.jpg", "4003951_12.jpg", "4003951_13.jpg", "4003951_14.jpg", "4003951_15.jpg" ]

[ { "from": "human", "value": "\nSuppose you are an expert in judging and evaluating the quality of AI-generated videos,\nplease watch the following frames of a given video and see the text prompt for generating the video,\nthen give scores from 5 different dimensions:\n(1) visual quality: the quality of the video in terms of clearness, resolution, brightness, and color\n(2) temporal consistency, both the consistency of objects or humans and the smoothness of motion or movements\n(3) dynamic degree, the degree of dynamic changes\n(4) text-to-video alignment, the alignment between the text prompt and the video content\n(5) factual consistency, the consistency of the video content with the common-sense and factual knowledge\n\nfor each dimension, output a float number from 1.0 to 4.0,\nthe higher the number is, the better the video performs in that sub-score, \nthe lowest 1.0 means Bad, the highest 4.0 means Perfect/Real (the video is like a real video)\nHere is an output example:\nvisual quality: 3.2\ntemporal consistency: 2.7\ndynamic degree: 4.0\ntext-to-video alignment: 2.3\nfactual consistency: 1.8\n\nFor this video, the text prompt is \"A luminous flying saucer flies away from the roof of the house.\",\nall the frames of video are as follows:\n\n<image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> " }, { "from": "gpt", "value": "visual quality: 1\n temporal consistency: 3\n dynamic degree: 3\n text-to-video alignment: 0\n factual consistency: 1\n" } ]

[ "6004280_00.jpg", "6004280_01.jpg", "6004280_02.jpg", "6004280_03.jpg", "6004280_04.jpg", "6004280_05.jpg", "6004280_06.jpg", "6004280_07.jpg", "6004280_08.jpg", "6004280_09.jpg", "6004280_10.jpg", "6004280_11.jpg", "6004280_12.jpg", "6004280_13.jpg", "6004280_14.jpg", "6004280_15.jpg" ]

[ { "from": "human", "value": "\nSuppose you are an expert in judging and evaluating the quality of AI-generated videos,\nplease watch the following frames of a given video and see the text prompt for generating the video,\nthen give scores from 5 different dimensions:\n(1) visual quality: the quality of the video in terms of clearness, resolution, brightness, and color\n(2) temporal consistency, both the consistency of objects or humans and the smoothness of motion or movements\n(3) dynamic degree, the degree of dynamic changes\n(4) text-to-video alignment, the alignment between the text prompt and the video content\n(5) factual consistency, the consistency of the video content with the common-sense and factual knowledge\n\nfor each dimension, output a float number from 1.0 to 4.0,\nthe higher the number is, the better the video performs in that sub-score, \nthe lowest 1.0 means Bad, the highest 4.0 means Perfect/Real (the video is like a real video)\nHere is an output example:\nvisual quality: 3.2\ntemporal consistency: 2.7\ndynamic degree: 4.0\ntext-to-video alignment: 2.3\nfactual consistency: 1.8\n\nFor this video, the text prompt is \"2 people sit at a table and shout\",\nall the frames of video are as follows:\n\n<image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> " }, { "from": "gpt", "value": "visual quality: 1\n temporal consistency: 1\n dynamic degree: 3\n text-to-video alignment: 0\n factual consistency: 1\n" } ]

[ "2001593_00.jpg", "2001593_01.jpg", "2001593_02.jpg", "2001593_03.jpg", "2001593_04.jpg", "2001593_05.jpg", "2001593_06.jpg", "2001593_07.jpg", "2001593_08.jpg", "2001593_09.jpg", "2001593_10.jpg", "2001593_11.jpg", "2001593_12.jpg", "2001593_13.jpg", "2001593_14.jpg", "2001593_15.jpg" ]

[ { "from": "human", "value": "\nSuppose you are an expert in judging and evaluating the quality of AI-generated videos,\nplease watch the following frames of a given video and see the text prompt for generating the video,\nthen give scores from 5 different dimensions:\n(1) visual quality: the quality of the video in terms of clearness, resolution, brightness, and color\n(2) temporal consistency, both the consistency of objects or humans and the smoothness of motion or movements\n(3) dynamic degree, the degree of dynamic changes\n(4) text-to-video alignment, the alignment between the text prompt and the video content\n(5) factual consistency, the consistency of the video content with the common-sense and factual knowledge\n\nfor each dimension, output a float number from 1.0 to 4.0,\nthe higher the number is, the better the video performs in that sub-score, \nthe lowest 1.0 means Bad, the highest 4.0 means Perfect/Real (the video is like a real video)\nHere is an output example:\nvisual quality: 3.2\ntemporal consistency: 2.7\ndynamic degree: 4.0\ntext-to-video alignment: 2.3\nfactual consistency: 1.8\n\nFor this video, the text prompt is \"Create a disney pixar 3d style the image of a Caucasian girl dressed in harlekin outfit in blue and red colors, with yellow pompom on the two ends of the hat and three yellow pompoms in place of the buttons of the outfit, big eyes, her pupils are like small circus tents seen from above. Her appearance is stocky and her head is slightly larger than her body, her hair is brown, short, and she has bangs, a youthful and beautiful face, hands with white gloves like the character Mickey Mouse.\",\nall the frames of video are as follows:\n\n<image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> " }, { "from": "gpt", "value": "visual quality: 2\n temporal consistency: 2\n dynamic degree: 3\n text-to-video alignment: 0\n factual consistency: 1\n" } ]

[ "8001013_00.jpg", "8001013_01.jpg", "8001013_02.jpg", "8001013_03.jpg", "8001013_04.jpg", "8001013_05.jpg", "8001013_06.jpg", "8001013_07.jpg", "8001013_08.jpg", "8001013_09.jpg", "8001013_10.jpg", "8001013_11.jpg", "8001013_12.jpg", "8001013_13.jpg", "8001013_14.jpg", "8001013_15.jpg", "8001013_16.jpg", "8001013_17.jpg", "8001013_18.jpg", "8001013_19.jpg", "8001013_20.jpg", "8001013_21.jpg", "8001013_22.jpg", "8001013_23.jpg" ]

[ { "from": "human", "value": "\nSuppose you are an expert in judging and evaluating the quality of AI-generated videos,\nplease watch the following frames of a given video and see the text prompt for generating the video,\nthen give scores from 5 different dimensions:\n(1) visual quality: the quality of the video in terms of clearness, resolution, brightness, and color\n(2) temporal consistency, both the consistency of objects or humans and the smoothness of motion or movements\n(3) dynamic degree, the degree of dynamic changes\n(4) text-to-video alignment, the alignment between the text prompt and the video content\n(5) factual consistency, the consistency of the video content with the common-sense and factual knowledge\n\nfor each dimension, output a float number from 1.0 to 4.0,\nthe higher the number is, the better the video performs in that sub-score, \nthe lowest 1.0 means Bad, the highest 4.0 means Perfect/Real (the video is like a real video)\nHere is an output example:\nvisual quality: 3.2\ntemporal consistency: 2.7\ndynamic degree: 4.0\ntext-to-video alignment: 2.3\nfactual consistency: 1.8\n\nFor this video, the text prompt is \"illustrate plenty of squirrels enjoy the wonders of the magical garden, spreading joy and happiness throughout the forest. Disney style\",\nall the frames of video are as follows:\n\n<image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> " }, { "from": "gpt", "value": "visual quality: 1\n temporal consistency: 2\n dynamic degree: 3\n text-to-video alignment: 0\n factual consistency: 1\n" } ]

[ "2002834_00.jpg", "2002834_01.jpg", "2002834_02.jpg", "2002834_03.jpg", "2002834_04.jpg", "2002834_05.jpg", "2002834_06.jpg", "2002834_07.jpg", "2002834_08.jpg", "2002834_09.jpg", "2002834_10.jpg", "2002834_11.jpg", "2002834_12.jpg", "2002834_13.jpg", "2002834_14.jpg", "2002834_15.jpg" ]

[ { "from": "human", "value": "\nSuppose you are an expert in judging and evaluating the quality of AI-generated videos,\nplease watch the following frames of a given video and see the text prompt for generating the video,\nthen give scores from 5 different dimensions:\n(1) visual quality: the quality of the video in terms of clearness, resolution, brightness, and color\n(2) temporal consistency, both the consistency of objects or humans and the smoothness of motion or movements\n(3) dynamic degree, the degree of dynamic changes\n(4) text-to-video alignment, the alignment between the text prompt and the video content\n(5) factual consistency, the consistency of the video content with the common-sense and factual knowledge\n\nfor each dimension, output a float number from 1.0 to 4.0,\nthe higher the number is, the better the video performs in that sub-score, \nthe lowest 1.0 means Bad, the highest 4.0 means Perfect/Real (the video is like a real video)\nHere is an output example:\nvisual quality: 3.2\ntemporal consistency: 2.7\ndynamic degree: 4.0\ntext-to-video alignment: 2.3\nfactual consistency: 1.8\n\nFor this video, the text prompt is \"pixar style animated shopping mall full of holiday products\",\nall the frames of video are as follows:\n\n<image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> " }, { "from": "gpt", "value": "visual quality: 2\n temporal consistency: 2\n dynamic degree: 3\n text-to-video alignment: 0\n factual consistency: 3\n" } ]

[ "2002088_00.jpg", "2002088_01.jpg", "2002088_02.jpg", "2002088_03.jpg", "2002088_04.jpg", "2002088_05.jpg", "2002088_06.jpg", "2002088_07.jpg", "2002088_08.jpg", "2002088_09.jpg", "2002088_10.jpg", "2002088_11.jpg", "2002088_12.jpg", "2002088_13.jpg", "2002088_14.jpg", "2002088_15.jpg" ]

[ { "from": "human", "value": "\nSuppose you are an expert in judging and evaluating the quality of AI-generated videos,\nplease watch the following frames of a given video and see the text prompt for generating the video,\nthen give scores from 5 different dimensions:\n(1) visual quality: the quality of the video in terms of clearness, resolution, brightness, and color\n(2) temporal consistency, both the consistency of objects or humans and the smoothness of motion or movements\n(3) dynamic degree, the degree of dynamic changes\n(4) text-to-video alignment, the alignment between the text prompt and the video content\n(5) factual consistency, the consistency of the video content with the common-sense and factual knowledge\n\nfor each dimension, output a float number from 1.0 to 4.0,\nthe higher the number is, the better the video performs in that sub-score, \nthe lowest 1.0 means Bad, the highest 4.0 means Perfect/Real (the video is like a real video)\nHere is an output example:\nvisual quality: 3.2\ntemporal consistency: 2.7\ndynamic degree: 4.0\ntext-to-video alignment: 2.3\nfactual consistency: 1.8\n\nFor this video, the text prompt is \"a girl with her pet walking in the jungle\",\nall the frames of video are as follows:\n\n<image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> " }, { "from": "gpt", "value": "visual quality: 3\n temporal consistency: 3\n dynamic degree: 3\n text-to-video alignment: 0\n factual consistency: 3\n" } ]

[ "3003560_00.jpg", "3003560_01.jpg", "3003560_02.jpg", "3003560_03.jpg", "3003560_04.jpg", "3003560_05.jpg", "3003560_06.jpg", "3003560_07.jpg", "3003560_08.jpg", "3003560_09.jpg", "3003560_10.jpg", "3003560_11.jpg", "3003560_12.jpg", "3003560_13.jpg", "3003560_14.jpg", "3003560_15.jpg" ]

[ { "from": "human", "value": "\nSuppose you are an expert in judging and evaluating the quality of AI-generated videos,\nplease watch the following frames of a given video and see the text prompt for generating the video,\nthen give scores from 5 different dimensions:\n(1) visual quality: the quality of the video in terms of clearness, resolution, brightness, and color\n(2) temporal consistency, both the consistency of objects or humans and the smoothness of motion or movements\n(3) dynamic degree, the degree of dynamic changes\n(4) text-to-video alignment, the alignment between the text prompt and the video content\n(5) factual consistency, the consistency of the video content with the common-sense and factual knowledge\n\nfor each dimension, output a float number from 1.0 to 4.0,\nthe higher the number is, the better the video performs in that sub-score, \nthe lowest 1.0 means Bad, the highest 4.0 means Perfect/Real (the video is like a real video)\nHere is an output example:\nvisual quality: 3.2\ntemporal consistency: 2.7\ndynamic degree: 4.0\ntext-to-video alignment: 2.3\nfactual consistency: 1.8\n\nFor this video, the text prompt is \"an elephant on a deserted island, stuck in loose rocks, and monkeys help him get out Message: elephant (Font: MODERN)\",\nall the frames of video are as follows:\n\n<image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> " }, { "from": "gpt", "value": "visual quality: 2\n temporal consistency: 3\n dynamic degree: 3\n text-to-video alignment: 0\n factual consistency: 2\n" } ]

[ "0001202_00.jpg", "0001202_01.jpg", "0001202_02.jpg", "0001202_03.jpg", "0001202_04.jpg", "0001202_05.jpg", "0001202_06.jpg", "0001202_07.jpg", "0001202_08.jpg", "0001202_09.jpg", "0001202_10.jpg", "0001202_11.jpg", "0001202_12.jpg", "0001202_13.jpg", "0001202_14.jpg", "0001202_15.jpg", "0001202_16.jpg", "0001202_17.jpg", "0001202_18.jpg", "0001202_19.jpg", "0001202_20.jpg", "0001202_21.jpg", "0001202_22.jpg", "0001202_23.jpg" ]

[ { "from": "human", "value": "\nSuppose you are an expert in judging and evaluating the quality of AI-generated videos,\nplease watch the following frames of a given video and see the text prompt for generating the video,\nthen give scores from 5 different dimensions:\n(1) visual quality: the quality of the video in terms of clearness, resolution, brightness, and color\n(2) temporal consistency, both the consistency of objects or humans and the smoothness of motion or movements\n(3) dynamic degree, the degree of dynamic changes\n(4) text-to-video alignment, the alignment between the text prompt and the video content\n(5) factual consistency, the consistency of the video content with the common-sense and factual knowledge\n\nfor each dimension, output a float number from 1.0 to 4.0,\nthe higher the number is, the better the video performs in that sub-score, \nthe lowest 1.0 means Bad, the highest 4.0 means Perfect/Real (the video is like a real video)\nHere is an output example:\nvisual quality: 3.2\ntemporal consistency: 2.7\ndynamic degree: 4.0\ntext-to-video alignment: 2.3\nfactual consistency: 1.8\n\nFor this video, the text prompt is \"Ravi a 20 years old men, sitting under a banyan tree, discussing his dreams with his wise grandfather. The setting sun casting long shadows as Ravi makes up his mind.\",\nall the frames of video are as follows:\n\n<image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> " }, { "from": "gpt", "value": "visual quality: 3\n temporal consistency: 3\n dynamic degree: 2\n text-to-video alignment: 0\n factual consistency: 3\n" } ]

[ "6002027_00.jpg", "6002027_01.jpg", "6002027_02.jpg", "6002027_03.jpg", "6002027_04.jpg", "6002027_05.jpg", "6002027_06.jpg", "6002027_07.jpg", "6002027_08.jpg", "6002027_09.jpg", "6002027_10.jpg", "6002027_11.jpg", "6002027_12.jpg", "6002027_13.jpg", "6002027_14.jpg", "6002027_15.jpg" ]

[ { "from": "human", "value": "\nSuppose you are an expert in judging and evaluating the quality of AI-generated videos,\nplease watch the following frames of a given video and see the text prompt for generating the video,\nthen give scores from 5 different dimensions:\n(1) visual quality: the quality of the video in terms of clearness, resolution, brightness, and color\n(2) temporal consistency, both the consistency of objects or humans and the smoothness of motion or movements\n(3) dynamic degree, the degree of dynamic changes\n(4) text-to-video alignment, the alignment between the text prompt and the video content\n(5) factual consistency, the consistency of the video content with the common-sense and factual knowledge\n\nfor each dimension, output a float number from 1.0 to 4.0,\nthe higher the number is, the better the video performs in that sub-score, \nthe lowest 1.0 means Bad, the highest 4.0 means Perfect/Real (the video is like a real video)\nHere is an output example:\nvisual quality: 3.2\ntemporal consistency: 2.7\ndynamic degree: 4.0\ntext-to-video alignment: 2.3\nfactual consistency: 1.8\n\nFor this video, the text prompt is \"a rapid sequence of arduino connected to other sensors\",\nall the frames of video are as follows:\n\n<image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> " }, { "from": "gpt", "value": "visual quality: 1\n temporal consistency: 1\n dynamic degree: 3\n text-to-video alignment: 0\n factual consistency: 2\n" } ]

[ "6001938_00.jpg", "6001938_01.jpg", "6001938_02.jpg", "6001938_03.jpg", "6001938_04.jpg", "6001938_05.jpg", "6001938_06.jpg", "6001938_07.jpg", "6001938_08.jpg", "6001938_09.jpg", "6001938_10.jpg", "6001938_11.jpg", "6001938_12.jpg", "6001938_13.jpg", "6001938_14.jpg", "6001938_15.jpg" ]

[ { "from": "human", "value": "\nSuppose you are an expert in judging and evaluating the quality of AI-generated videos,\nplease watch the following frames of a given video and see the text prompt for generating the video,\nthen give scores from 5 different dimensions:\n(1) visual quality: the quality of the video in terms of clearness, resolution, brightness, and color\n(2) temporal consistency, both the consistency of objects or humans and the smoothness of motion or movements\n(3) dynamic degree, the degree of dynamic changes\n(4) text-to-video alignment, the alignment between the text prompt and the video content\n(5) factual consistency, the consistency of the video content with the common-sense and factual knowledge\n\nfor each dimension, output a float number from 1.0 to 4.0,\nthe higher the number is, the better the video performs in that sub-score, \nthe lowest 1.0 means Bad, the highest 4.0 means Perfect/Real (the video is like a real video)\nHere is an output example:\nvisual quality: 3.2\ntemporal consistency: 2.7\ndynamic degree: 4.0\ntext-to-video alignment: 2.3\nfactual consistency: 1.8\n\nFor this video, the text prompt is \"a girl anime with red eyes, cherry blossom petals arond her\",\nall the frames of video are as follows:\n\n<image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> " }, { "from": "gpt", "value": "visual quality: 1\n temporal consistency: 2\n dynamic degree: 3\n text-to-video alignment: 0\n factual consistency: 1\n" } ]

[ "7002023_00.jpg", "7002023_01.jpg", "7002023_02.jpg", "7002023_03.jpg", "7002023_04.jpg", "7002023_05.jpg", "7002023_06.jpg", "7002023_07.jpg" ]

[ { "from": "human", "value": "\nSuppose you are an expert in judging and evaluating the quality of AI-generated videos,\nplease watch the following frames of a given video and see the text prompt for generating the video,\nthen give scores from 5 different dimensions:\n(1) visual quality: the quality of the video in terms of clearness, resolution, brightness, and color\n(2) temporal consistency, both the consistency of objects or humans and the smoothness of motion or movements\n(3) dynamic degree, the degree of dynamic changes\n(4) text-to-video alignment, the alignment between the text prompt and the video content\n(5) factual consistency, the consistency of the video content with the common-sense and factual knowledge\n\nfor each dimension, output a float number from 1.0 to 4.0,\nthe higher the number is, the better the video performs in that sub-score, \nthe lowest 1.0 means Bad, the highest 4.0 means Perfect/Real (the video is like a real video)\nHere is an output example:\nvisual quality: 3.2\ntemporal consistency: 2.7\ndynamic degree: 4.0\ntext-to-video alignment: 2.3\nfactual consistency: 1.8\n\nFor this video, the text prompt is \"An array of 1000 solar cells paved on a grassland take the sunshine and convert it into electricity, which is utilized for water electrolysis\",\nall the frames of video are as follows:\n\n<image> <image> <image> <image> <image> <image> <image> <image> " }, { "from": "gpt", "value": "visual quality: 2\n temporal consistency: 2\n dynamic degree: 3\n text-to-video alignment: 0\n factual consistency: 2\n" } ]

[ "4000096_00.jpg", "4000096_01.jpg", "4000096_02.jpg", "4000096_03.jpg", "4000096_04.jpg", "4000096_05.jpg", "4000096_06.jpg", "4000096_07.jpg", "4000096_08.jpg", "4000096_09.jpg", "4000096_10.jpg", "4000096_11.jpg", "4000096_12.jpg", "4000096_13.jpg", "4000096_14.jpg", "4000096_15.jpg" ]

[ { "from": "human", "value": "\nSuppose you are an expert in judging and evaluating the quality of AI-generated videos,\nplease watch the following frames of a given video and see the text prompt for generating the video,\nthen give scores from 5 different dimensions:\n(1) visual quality: the quality of the video in terms of clearness, resolution, brightness, and color\n(2) temporal consistency, both the consistency of objects or humans and the smoothness of motion or movements\n(3) dynamic degree, the degree of dynamic changes\n(4) text-to-video alignment, the alignment between the text prompt and the video content\n(5) factual consistency, the consistency of the video content with the common-sense and factual knowledge\n\nfor each dimension, output a float number from 1.0 to 4.0,\nthe higher the number is, the better the video performs in that sub-score, \nthe lowest 1.0 means Bad, the highest 4.0 means Perfect/Real (the video is like a real video)\nHere is an output example:\nvisual quality: 3.2\ntemporal consistency: 2.7\ndynamic degree: 4.0\ntext-to-video alignment: 2.3\nfactual consistency: 1.8\n\nFor this video, the text prompt is \"a woman waits for somebody near her house. She looks on the road. She wears dark modest dress. seed5622540067021185986\",\nall the frames of video are as follows:\n\n<image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> " }, { "from": "gpt", "value": "visual quality: 3\n temporal consistency: 2\n dynamic degree: 2\n text-to-video alignment: 0\n factual consistency: 3\n" } ]

[ "8003697_00.jpg", "8003697_01.jpg", "8003697_02.jpg", "8003697_03.jpg", "8003697_04.jpg", "8003697_05.jpg", "8003697_06.jpg", "8003697_07.jpg", "8003697_08.jpg", "8003697_09.jpg", "8003697_10.jpg", "8003697_11.jpg", "8003697_12.jpg", "8003697_13.jpg", "8003697_14.jpg", "8003697_15.jpg", "8003697_16.jpg", "8003697_17.jpg", "8003697_18.jpg", "8003697_19.jpg", "8003697_20.jpg", "8003697_21.jpg", "8003697_22.jpg", "8003697_23.jpg" ]

[ { "from": "human", "value": "\nSuppose you are an expert in judging and evaluating the quality of AI-generated videos,\nplease watch the following frames of a given video and see the text prompt for generating the video,\nthen give scores from 5 different dimensions:\n(1) visual quality: the quality of the video in terms of clearness, resolution, brightness, and color\n(2) temporal consistency, both the consistency of objects or humans and the smoothness of motion or movements\n(3) dynamic degree, the degree of dynamic changes\n(4) text-to-video alignment, the alignment between the text prompt and the video content\n(5) factual consistency, the consistency of the video content with the common-sense and factual knowledge\n\nfor each dimension, output a float number from 1.0 to 4.0,\nthe higher the number is, the better the video performs in that sub-score, \nthe lowest 1.0 means Bad, the highest 4.0 means Perfect/Real (the video is like a real video)\nHere is an output example:\nvisual quality: 3.2\ntemporal consistency: 2.7\ndynamic degree: 4.0\ntext-to-video alignment: 2.3\nfactual consistency: 1.8\n\nFor this video, the text prompt is \"The butterfly continued to crawl, unable to fly\",\nall the frames of video are as follows:\n\n<image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> " }, { "from": "gpt", "value": "visual quality: 1\n temporal consistency: 1\n dynamic degree: 3\n text-to-video alignment: 0\n factual consistency: 1\n" } ]

[ "4000085_00.jpg", "4000085_01.jpg", "4000085_02.jpg", "4000085_03.jpg", "4000085_04.jpg", "4000085_05.jpg", "4000085_06.jpg", "4000085_07.jpg", "4000085_08.jpg", "4000085_09.jpg", "4000085_10.jpg", "4000085_11.jpg", "4000085_12.jpg", "4000085_13.jpg", "4000085_14.jpg", "4000085_15.jpg" ]

[ { "from": "human", "value": "\nSuppose you are an expert in judging and evaluating the quality of AI-generated videos,\nplease watch the following frames of a given video and see the text prompt for generating the video,\nthen give scores from 5 different dimensions:\n(1) visual quality: the quality of the video in terms of clearness, resolution, brightness, and color\n(2) temporal consistency, both the consistency of objects or humans and the smoothness of motion or movements\n(3) dynamic degree, the degree of dynamic changes\n(4) text-to-video alignment, the alignment between the text prompt and the video content\n(5) factual consistency, the consistency of the video content with the common-sense and factual knowledge\n\nfor each dimension, output a float number from 1.0 to 4.0,\nthe higher the number is, the better the video performs in that sub-score, \nthe lowest 1.0 means Bad, the highest 4.0 means Perfect/Real (the video is like a real video)\nHere is an output example:\nvisual quality: 3.2\ntemporal consistency: 2.7\ndynamic degree: 4.0\ntext-to-video alignment: 2.3\nfactual consistency: 1.8\n\nFor this video, the text prompt is \"girl, blue hair, dancing in the disco\",\nall the frames of video are as follows:\n\n<image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> " }, { "from": "gpt", "value": "visual quality: 2\n temporal consistency: 3\n dynamic degree: 3\n text-to-video alignment: 0\n factual consistency: 3\n" } ]

[ "7003453_00.jpg", "7003453_01.jpg", "7003453_02.jpg", "7003453_03.jpg", "7003453_04.jpg", "7003453_05.jpg", "7003453_06.jpg", "7003453_07.jpg" ]

[ { "from": "human", "value": "\nSuppose you are an expert in judging and evaluating the quality of AI-generated videos,\nplease watch the following frames of a given video and see the text prompt for generating the video,\nthen give scores from 5 different dimensions:\n(1) visual quality: the quality of the video in terms of clearness, resolution, brightness, and color\n(2) temporal consistency, both the consistency of objects or humans and the smoothness of motion or movements\n(3) dynamic degree, the degree of dynamic changes\n(4) text-to-video alignment, the alignment between the text prompt and the video content\n(5) factual consistency, the consistency of the video content with the common-sense and factual knowledge\n\nfor each dimension, output a float number from 1.0 to 4.0,\nthe higher the number is, the better the video performs in that sub-score, \nthe lowest 1.0 means Bad, the highest 4.0 means Perfect/Real (the video is like a real video)\nHere is an output example:\nvisual quality: 3.2\ntemporal consistency: 2.7\ndynamic degree: 4.0\ntext-to-video alignment: 2.3\nfactual consistency: 1.8\n\nFor this video, the text prompt is \"In a still frame, nestled in the Zen garden, a rustic teahouse featured tatami seating and a traditional charcoal brazier\",\nall the frames of video are as follows:\n\n<image> <image> <image> <image> <image> <image> <image> <image> " }, { "from": "gpt", "value": "visual quality: 3\n temporal consistency: 2\n dynamic degree: 3\n text-to-video alignment: 0\n factual consistency: 1\n" } ]

[ "3002841_00.jpg", "3002841_01.jpg", "3002841_02.jpg", "3002841_03.jpg", "3002841_04.jpg", "3002841_05.jpg", "3002841_06.jpg", "3002841_07.jpg", "3002841_08.jpg", "3002841_09.jpg", "3002841_10.jpg", "3002841_11.jpg", "3002841_12.jpg", "3002841_13.jpg", "3002841_14.jpg", "3002841_15.jpg" ]

[ { "from": "human", "value": "\nSuppose you are an expert in judging and evaluating the quality of AI-generated videos,\nplease watch the following frames of a given video and see the text prompt for generating the video,\nthen give scores from 5 different dimensions:\n(1) visual quality: the quality of the video in terms of clearness, resolution, brightness, and color\n(2) temporal consistency, both the consistency of objects or humans and the smoothness of motion or movements\n(3) dynamic degree, the degree of dynamic changes\n(4) text-to-video alignment, the alignment between the text prompt and the video content\n(5) factual consistency, the consistency of the video content with the common-sense and factual knowledge\n\nfor each dimension, output a float number from 1.0 to 4.0,\nthe higher the number is, the better the video performs in that sub-score, \nthe lowest 1.0 means Bad, the highest 4.0 means Perfect/Real (the video is like a real video)\nHere is an output example:\nvisual quality: 3.2\ntemporal consistency: 2.7\ndynamic degree: 4.0\ntext-to-video alignment: 2.3\nfactual consistency: 1.8\n\nFor this video, the text prompt is \"flying bicycles against the backdrop of a sunset in the mountains\",\nall the frames of video are as follows:\n\n<image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> " }, { "from": "gpt", "value": "visual quality: 2\n temporal consistency: 1\n dynamic degree: 3\n text-to-video alignment: 0\n factual consistency: 2\n" } ]

[ "4000504_00.jpg", "4000504_01.jpg", "4000504_02.jpg", "4000504_03.jpg", "4000504_04.jpg", "4000504_05.jpg", "4000504_06.jpg", "4000504_07.jpg", "4000504_08.jpg", "4000504_09.jpg", "4000504_10.jpg", "4000504_11.jpg", "4000504_12.jpg", "4000504_13.jpg", "4000504_14.jpg", "4000504_15.jpg" ]

[ { "from": "human", "value": "\nSuppose you are an expert in judging and evaluating the quality of AI-generated videos,\nplease watch the following frames of a given video and see the text prompt for generating the video,\nthen give scores from 5 different dimensions:\n(1) visual quality: the quality of the video in terms of clearness, resolution, brightness, and color\n(2) temporal consistency, both the consistency of objects or humans and the smoothness of motion or movements\n(3) dynamic degree, the degree of dynamic changes\n(4) text-to-video alignment, the alignment between the text prompt and the video content\n(5) factual consistency, the consistency of the video content with the common-sense and factual knowledge\n\nfor each dimension, output a float number from 1.0 to 4.0,\nthe higher the number is, the better the video performs in that sub-score, \nthe lowest 1.0 means Bad, the highest 4.0 means Perfect/Real (the video is like a real video)\nHere is an output example:\nvisual quality: 3.2\ntemporal consistency: 2.7\ndynamic degree: 4.0\ntext-to-video alignment: 2.3\nfactual consistency: 1.8\n\nFor this video, the text prompt is \"3d cartoon in a cave, one pirate pushed other down in the lava\",\nall the frames of video are as follows:\n\n<image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> " }, { "from": "gpt", "value": "visual quality: 2\n temporal consistency: 2\n dynamic degree: 3\n text-to-video alignment: 0\n factual consistency: 3\n" } ]

[ "1001055_00.jpg", "1001055_01.jpg", "1001055_02.jpg", "1001055_03.jpg", "1001055_04.jpg", "1001055_05.jpg", "1001055_06.jpg", "1001055_07.jpg", "1001055_08.jpg", "1001055_09.jpg", "1001055_10.jpg", "1001055_11.jpg", "1001055_12.jpg", "1001055_13.jpg", "1001055_14.jpg", "1001055_15.jpg" ]

[ { "from": "human", "value": "\nSuppose you are an expert in judging and evaluating the quality of AI-generated videos,\nplease watch the following frames of a given video and see the text prompt for generating the video,\nthen give scores from 5 different dimensions:\n(1) visual quality: the quality of the video in terms of clearness, resolution, brightness, and color\n(2) temporal consistency, both the consistency of objects or humans and the smoothness of motion or movements\n(3) dynamic degree, the degree of dynamic changes\n(4) text-to-video alignment, the alignment between the text prompt and the video content\n(5) factual consistency, the consistency of the video content with the common-sense and factual knowledge\n\nfor each dimension, output a float number from 1.0 to 4.0,\nthe higher the number is, the better the video performs in that sub-score, \nthe lowest 1.0 means Bad, the highest 4.0 means Perfect/Real (the video is like a real video)\nHere is an output example:\nvisual quality: 3.2\ntemporal consistency: 2.7\ndynamic degree: 4.0\ntext-to-video alignment: 2.3\nfactual consistency: 1.8\n\nFor this video, the text prompt is \"In a medieval kingdom, a young squire stumbles upon a portal to a futuristic world\",\nall the frames of video are as follows:\n\n<image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> " }, { "from": "gpt", "value": "visual quality: 1\n temporal consistency: 2\n dynamic degree: 3\n text-to-video alignment: 0\n factual consistency: 1\n" } ]

[ "4000289_00.jpg", "4000289_01.jpg", "4000289_02.jpg", "4000289_03.jpg", "4000289_04.jpg", "4000289_05.jpg", "4000289_06.jpg", "4000289_07.jpg", "4000289_08.jpg", "4000289_09.jpg", "4000289_10.jpg", "4000289_11.jpg", "4000289_12.jpg", "4000289_13.jpg", "4000289_14.jpg", "4000289_15.jpg" ]

[ { "from": "human", "value": "\nSuppose you are an expert in judging and evaluating the quality of AI-generated videos,\nplease watch the following frames of a given video and see the text prompt for generating the video,\nthen give scores from 5 different dimensions:\n(1) visual quality: the quality of the video in terms of clearness, resolution, brightness, and color\n(2) temporal consistency, both the consistency of objects or humans and the smoothness of motion or movements\n(3) dynamic degree, the degree of dynamic changes\n(4) text-to-video alignment, the alignment between the text prompt and the video content\n(5) factual consistency, the consistency of the video content with the common-sense and factual knowledge\n\nfor each dimension, output a float number from 1.0 to 4.0,\nthe higher the number is, the better the video performs in that sub-score, \nthe lowest 1.0 means Bad, the highest 4.0 means Perfect/Real (the video is like a real video)\nHere is an output example:\nvisual quality: 3.2\ntemporal consistency: 2.7\ndynamic degree: 4.0\ntext-to-video alignment: 2.3\nfactual consistency: 1.8\n\nFor this video, the text prompt is \"Beautiful tango dancers, wide shot, 8k, realistic, cinematic\",\nall the frames of video are as follows:\n\n<image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> " }, { "from": "gpt", "value": "visual quality: 3\n temporal consistency: 3\n dynamic degree: 3\n text-to-video alignment: 0\n factual consistency: 3\n" } ]

[ "3000081_00.jpg", "3000081_01.jpg", "3000081_02.jpg", "3000081_03.jpg", "3000081_04.jpg", "3000081_05.jpg", "3000081_06.jpg", "3000081_07.jpg", "3000081_08.jpg", "3000081_09.jpg", "3000081_10.jpg", "3000081_11.jpg", "3000081_12.jpg", "3000081_13.jpg", "3000081_14.jpg", "3000081_15.jpg" ]

[ { "from": "human", "value": "\nSuppose you are an expert in judging and evaluating the quality of AI-generated videos,\nplease watch the following frames of a given video and see the text prompt for generating the video,\nthen give scores from 5 different dimensions:\n(1) visual quality: the quality of the video in terms of clearness, resolution, brightness, and color\n(2) temporal consistency, both the consistency of objects or humans and the smoothness of motion or movements\n(3) dynamic degree, the degree of dynamic changes\n(4) text-to-video alignment, the alignment between the text prompt and the video content\n(5) factual consistency, the consistency of the video content with the common-sense and factual knowledge\n\nfor each dimension, output a float number from 1.0 to 4.0,\nthe higher the number is, the better the video performs in that sub-score, \nthe lowest 1.0 means Bad, the highest 4.0 means Perfect/Real (the video is like a real video)\nHere is an output example:\nvisual quality: 3.2\ntemporal consistency: 2.7\ndynamic degree: 4.0\ntext-to-video alignment: 2.3\nfactual consistency: 1.8\n\nFor this video, the text prompt is \"3D Animation a donkey drinking water at the lake with other animals of forest\",\nall the frames of video are as follows:\n\n<image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> " }, { "from": "gpt", "value": "visual quality: 2\n temporal consistency: 3\n dynamic degree: 1\n text-to-video alignment: 0\n factual consistency: 3\n" } ]

[ "6000251_00.jpg", "6000251_01.jpg", "6000251_02.jpg", "6000251_03.jpg", "6000251_04.jpg", "6000251_05.jpg", "6000251_06.jpg", "6000251_07.jpg", "6000251_08.jpg", "6000251_09.jpg", "6000251_10.jpg", "6000251_11.jpg", "6000251_12.jpg", "6000251_13.jpg", "6000251_14.jpg", "6000251_15.jpg" ]

[ { "from": "human", "value": "\nSuppose you are an expert in judging and evaluating the quality of AI-generated videos,\nplease watch the following frames of a given video and see the text prompt for generating the video,\nthen give scores from 5 different dimensions:\n(1) visual quality: the quality of the video in terms of clearness, resolution, brightness, and color\n(2) temporal consistency, both the consistency of objects or humans and the smoothness of motion or movements\n(3) dynamic degree, the degree of dynamic changes\n(4) text-to-video alignment, the alignment between the text prompt and the video content\n(5) factual consistency, the consistency of the video content with the common-sense and factual knowledge\n\nfor each dimension, output a float number from 1.0 to 4.0,\nthe higher the number is, the better the video performs in that sub-score, \nthe lowest 1.0 means Bad, the highest 4.0 means Perfect/Real (the video is like a real video)\nHere is an output example:\nvisual quality: 3.2\ntemporal consistency: 2.7\ndynamic degree: 4.0\ntext-to-video alignment: 2.3\nfactual consistency: 1.8\n\nFor this video, the text prompt is \"A montage of man and woman achieving their goals through hard work and dedication\",\nall the frames of video are as follows:\n\n<image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> " }, { "from": "gpt", "value": "visual quality: 1\n temporal consistency: 1\n dynamic degree: 3\n text-to-video alignment: 0\n factual consistency: 1\n" } ]

[ "3005504_00.jpg", "3005504_01.jpg", "3005504_02.jpg", "3005504_03.jpg", "3005504_04.jpg", "3005504_05.jpg", "3005504_06.jpg", "3005504_07.jpg", "3005504_08.jpg", "3005504_09.jpg", "3005504_10.jpg", "3005504_11.jpg", "3005504_12.jpg", "3005504_13.jpg", "3005504_14.jpg", "3005504_15.jpg" ]

[ { "from": "human", "value": "\nSuppose you are an expert in judging and evaluating the quality of AI-generated videos,\nplease watch the following frames of a given video and see the text prompt for generating the video,\nthen give scores from 5 different dimensions:\n(1) visual quality: the quality of the video in terms of clearness, resolution, brightness, and color\n(2) temporal consistency, both the consistency of objects or humans and the smoothness of motion or movements\n(3) dynamic degree, the degree of dynamic changes\n(4) text-to-video alignment, the alignment between the text prompt and the video content\n(5) factual consistency, the consistency of the video content with the common-sense and factual knowledge\n\nfor each dimension, output a float number from 1.0 to 4.0,\nthe higher the number is, the better the video performs in that sub-score, \nthe lowest 1.0 means Bad, the highest 4.0 means Perfect/Real (the video is like a real video)\nHere is an output example:\nvisual quality: 3.2\ntemporal consistency: 2.7\ndynamic degree: 4.0\ntext-to-video alignment: 2.3\nfactual consistency: 1.8\n\nFor this video, the text prompt is \"a hen spreads her wings and begins to fly, cinematic style\",\nall the frames of video are as follows:\n\n<image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> " }, { "from": "gpt", "value": "visual quality: 1\n temporal consistency: 2\n dynamic degree: 3\n text-to-video alignment: 0\n factual consistency: 2\n" } ]

[ "s000230_00.jpg", "s000230_01.jpg", "s000230_02.jpg", "s000230_03.jpg", "s000230_04.jpg", "s000230_05.jpg", "s000230_06.jpg", "s000230_07.jpg", "s000230_08.jpg", "s000230_09.jpg", "s000230_10.jpg", "s000230_11.jpg", "s000230_12.jpg", "s000230_13.jpg", "s000230_14.jpg", "s000230_15.jpg", "s000230_16.jpg", "s000230_17.jpg", "s000230_18.jpg", "s000230_19.jpg", "s000230_20.jpg", "s000230_21.jpg", "s000230_22.jpg", "s000230_23.jpg" ]

[ { "from": "human", "value": "\nSuppose you are an expert in judging and evaluating the quality of AI-generated videos,\nplease watch the following frames of a given video and see the text prompt for generating the video,\nthen give scores from 5 different dimensions:\n(1) visual quality: the quality of the video in terms of clearness, resolution, brightness, and color\n(2) temporal consistency, both the consistency of objects or humans and the smoothness of motion or movements\n(3) dynamic degree, the degree of dynamic changes\n(4) text-to-video alignment, the alignment between the text prompt and the video content\n(5) factual consistency, the consistency of the video content with the common-sense and factual knowledge\n\nfor each dimension, output a float number from 1.0 to 4.0,\nthe higher the number is, the better the video performs in that sub-score, \nthe lowest 1.0 means Bad, the highest 4.0 means Perfect/Real (the video is like a real video)\nHere is an output example:\nvisual quality: 3.2\ntemporal consistency: 2.7\ndynamic degree: 4.0\ntext-to-video alignment: 2.3\nfactual consistency: 1.8\n\nFor this video, the text prompt is \"Forest clearing ahead, the horseman nearing home in night in forest\",\nall the frames of video are as follows:\n\n<image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> " }, { "from": "gpt", "value": "visual quality: 4\n temporal consistency: 4\n dynamic degree: 4\n text-to-video alignment: 0\n factual consistency: 4\n" } ]

[ "4001047_00.jpg", "4001047_01.jpg", "4001047_02.jpg", "4001047_03.jpg", "4001047_04.jpg", "4001047_05.jpg", "4001047_06.jpg", "4001047_07.jpg", "4001047_08.jpg", "4001047_09.jpg", "4001047_10.jpg", "4001047_11.jpg", "4001047_12.jpg", "4001047_13.jpg", "4001047_14.jpg", "4001047_15.jpg" ]

[ { "from": "human", "value": "\nSuppose you are an expert in judging and evaluating the quality of AI-generated videos,\nplease watch the following frames of a given video and see the text prompt for generating the video,\nthen give scores from 5 different dimensions:\n(1) visual quality: the quality of the video in terms of clearness, resolution, brightness, and color\n(2) temporal consistency, both the consistency of objects or humans and the smoothness of motion or movements\n(3) dynamic degree, the degree of dynamic changes\n(4) text-to-video alignment, the alignment between the text prompt and the video content\n(5) factual consistency, the consistency of the video content with the common-sense and factual knowledge\n\nfor each dimension, output a float number from 1.0 to 4.0,\nthe higher the number is, the better the video performs in that sub-score, \nthe lowest 1.0 means Bad, the highest 4.0 means Perfect/Real (the video is like a real video)\nHere is an output example:\nvisual quality: 3.2\ntemporal consistency: 2.7\ndynamic degree: 4.0\ntext-to-video alignment: 2.3\nfactual consistency: 1.8\n\nFor this video, the text prompt is \"giant octopus alive, empty street background, black and white photograph, hyper realistic, motion 1\",\nall the frames of video are as follows:\n\n<image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> " }, { "from": "gpt", "value": "visual quality: 3\n temporal consistency: 2\n dynamic degree: 3\n text-to-video alignment: 0\n factual consistency: 3\n" } ]

[ "2001309_00.jpg", "2001309_01.jpg", "2001309_02.jpg", "2001309_03.jpg", "2001309_04.jpg", "2001309_05.jpg", "2001309_06.jpg", "2001309_07.jpg", "2001309_08.jpg", "2001309_09.jpg", "2001309_10.jpg", "2001309_11.jpg", "2001309_12.jpg", "2001309_13.jpg", "2001309_14.jpg", "2001309_15.jpg" ]

[ { "from": "human", "value": "\nSuppose you are an expert in judging and evaluating the quality of AI-generated videos,\nplease watch the following frames of a given video and see the text prompt for generating the video,\nthen give scores from 5 different dimensions:\n(1) visual quality: the quality of the video in terms of clearness, resolution, brightness, and color\n(2) temporal consistency, both the consistency of objects or humans and the smoothness of motion or movements\n(3) dynamic degree, the degree of dynamic changes\n(4) text-to-video alignment, the alignment between the text prompt and the video content\n(5) factual consistency, the consistency of the video content with the common-sense and factual knowledge\n\nfor each dimension, output a float number from 1.0 to 4.0,\nthe higher the number is, the better the video performs in that sub-score, \nthe lowest 1.0 means Bad, the highest 4.0 means Perfect/Real (the video is like a real video)\nHere is an output example:\nvisual quality: 3.2\ntemporal consistency: 2.7\ndynamic degree: 4.0\ntext-to-video alignment: 2.3\nfactual consistency: 1.8\n\nFor this video, the text prompt is \"1930s black and white film documentary about a dystopian society, in a rundown street where trash and buildings are boarded up and people seek refuge with barrels of fire in the streets\",\nall the frames of video are as follows:\n\n<image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> " }, { "from": "gpt", "value": "visual quality: 3\n temporal consistency: 2\n dynamic degree: 3\n text-to-video alignment: 0\n factual consistency: 3\n" } ]

[ "3002688_00.jpg", "3002688_01.jpg", "3002688_02.jpg", "3002688_03.jpg", "3002688_04.jpg", "3002688_05.jpg", "3002688_06.jpg", "3002688_07.jpg", "3002688_08.jpg", "3002688_09.jpg", "3002688_10.jpg", "3002688_11.jpg", "3002688_12.jpg", "3002688_13.jpg", "3002688_14.jpg", "3002688_15.jpg" ]

[ { "from": "human", "value": "\nSuppose you are an expert in judging and evaluating the quality of AI-generated videos,\nplease watch the following frames of a given video and see the text prompt for generating the video,\nthen give scores from 5 different dimensions:\n(1) visual quality: the quality of the video in terms of clearness, resolution, brightness, and color\n(2) temporal consistency, both the consistency of objects or humans and the smoothness of motion or movements\n(3) dynamic degree, the degree of dynamic changes\n(4) text-to-video alignment, the alignment between the text prompt and the video content\n(5) factual consistency, the consistency of the video content with the common-sense and factual knowledge\n\nfor each dimension, output a float number from 1.0 to 4.0,\nthe higher the number is, the better the video performs in that sub-score, \nthe lowest 1.0 means Bad, the highest 4.0 means Perfect/Real (the video is like a real video)\nHere is an output example:\nvisual quality: 3.2\ntemporal consistency: 2.7\ndynamic degree: 4.0\ntext-to-video alignment: 2.3\nfactual consistency: 1.8\n\nFor this video, the text prompt is \"a wolf viciously devouring a deer cinematic shot\",\nall the frames of video are as follows:\n\n<image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> " }, { "from": "gpt", "value": "visual quality: 1\n temporal consistency: 1\n dynamic degree: 3\n text-to-video alignment: 0\n factual consistency: 1\n" } ]

[ "2005385_00.jpg", "2005385_01.jpg", "2005385_02.jpg", "2005385_03.jpg", "2005385_04.jpg", "2005385_05.jpg", "2005385_06.jpg", "2005385_07.jpg", "2005385_08.jpg", "2005385_09.jpg", "2005385_10.jpg", "2005385_11.jpg", "2005385_12.jpg", "2005385_13.jpg", "2005385_14.jpg", "2005385_15.jpg" ]

[ { "from": "human", "value": "\nSuppose you are an expert in judging and evaluating the quality of AI-generated videos,\nplease watch the following frames of a given video and see the text prompt for generating the video,\nthen give scores from 5 different dimensions:\n(1) visual quality: the quality of the video in terms of clearness, resolution, brightness, and color\n(2) temporal consistency, both the consistency of objects or humans and the smoothness of motion or movements\n(3) dynamic degree, the degree of dynamic changes\n(4) text-to-video alignment, the alignment between the text prompt and the video content\n(5) factual consistency, the consistency of the video content with the common-sense and factual knowledge\n\nfor each dimension, output a float number from 1.0 to 4.0,\nthe higher the number is, the better the video performs in that sub-score, \nthe lowest 1.0 means Bad, the highest 4.0 means Perfect/Real (the video is like a real video)\nHere is an output example:\nvisual quality: 3.2\ntemporal consistency: 2.7\ndynamic degree: 4.0\ntext-to-video alignment: 2.3\nfactual consistency: 1.8\n\nFor this video, the text prompt is \"girl listening to music with headphone, lofi art\",\nall the frames of video are as follows:\n\n<image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> " }, { "from": "gpt", "value": "visual quality: 3\n temporal consistency: 3\n dynamic degree: 3\n text-to-video alignment: 0\n factual consistency: 3\n" } ]

[ "0003907_00.jpg", "0003907_01.jpg", "0003907_02.jpg", "0003907_03.jpg", "0003907_04.jpg", "0003907_05.jpg", "0003907_06.jpg", "0003907_07.jpg", "0003907_08.jpg", "0003907_09.jpg", "0003907_10.jpg", "0003907_11.jpg", "0003907_12.jpg", "0003907_13.jpg", "0003907_14.jpg", "0003907_15.jpg", "0003907_16.jpg", "0003907_17.jpg", "0003907_18.jpg", "0003907_19.jpg", "0003907_20.jpg", "0003907_21.jpg", "0003907_22.jpg", "0003907_23.jpg" ]

[ { "from": "human", "value": "\nSuppose you are an expert in judging and evaluating the quality of AI-generated videos,\nplease watch the following frames of a given video and see the text prompt for generating the video,\nthen give scores from 5 different dimensions:\n(1) visual quality: the quality of the video in terms of clearness, resolution, brightness, and color\n(2) temporal consistency, both the consistency of objects or humans and the smoothness of motion or movements\n(3) dynamic degree, the degree of dynamic changes\n(4) text-to-video alignment, the alignment between the text prompt and the video content\n(5) factual consistency, the consistency of the video content with the common-sense and factual knowledge\n\nfor each dimension, output a float number from 1.0 to 4.0,\nthe higher the number is, the better the video performs in that sub-score, \nthe lowest 1.0 means Bad, the highest 4.0 means Perfect/Real (the video is like a real video)\nHere is an output example:\nvisual quality: 3.2\ntemporal consistency: 2.7\ndynamic degree: 4.0\ntext-to-video alignment: 2.3\nfactual consistency: 1.8\n\nFor this video, the text prompt is \"Times square with cars flying and futurist building\",\nall the frames of video are as follows:\n\n<image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> " }, { "from": "gpt", "value": "visual quality: 3\n temporal consistency: 3\n dynamic degree: 1\n text-to-video alignment: 0\n factual consistency: 2\n" } ]

[ "1000085_00.jpg", "1000085_01.jpg", "1000085_02.jpg", "1000085_03.jpg", "1000085_04.jpg", "1000085_05.jpg", "1000085_06.jpg", "1000085_07.jpg", "1000085_08.jpg", "1000085_09.jpg", "1000085_10.jpg", "1000085_11.jpg", "1000085_12.jpg", "1000085_13.jpg", "1000085_14.jpg", "1000085_15.jpg" ]

[ { "from": "human", "value": "\nSuppose you are an expert in judging and evaluating the quality of AI-generated videos,\nplease watch the following frames of a given video and see the text prompt for generating the video,\nthen give scores from 5 different dimensions:\n(1) visual quality: the quality of the video in terms of clearness, resolution, brightness, and color\n(2) temporal consistency, both the consistency of objects or humans and the smoothness of motion or movements\n(3) dynamic degree, the degree of dynamic changes\n(4) text-to-video alignment, the alignment between the text prompt and the video content\n(5) factual consistency, the consistency of the video content with the common-sense and factual knowledge\n\nfor each dimension, output a float number from 1.0 to 4.0,\nthe higher the number is, the better the video performs in that sub-score, \nthe lowest 1.0 means Bad, the highest 4.0 means Perfect/Real (the video is like a real video)\nHere is an output example:\nvisual quality: 3.2\ntemporal consistency: 2.7\ndynamic degree: 4.0\ntext-to-video alignment: 2.3\nfactual consistency: 1.8\n\nFor this video, the text prompt is \"Hulk vs Godzilla with cartoon style\",\nall the frames of video are as follows:\n\n<image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> " }, { "from": "gpt", "value": "visual quality: 1\n temporal consistency: 1\n dynamic degree: 3\n text-to-video alignment: 0\n factual consistency: 1\n" } ]

[ "7000787_00.jpg", "7000787_01.jpg", "7000787_02.jpg", "7000787_03.jpg", "7000787_04.jpg", "7000787_05.jpg", "7000787_06.jpg", "7000787_07.jpg" ]

[ { "from": "human", "value": "\nSuppose you are an expert in judging and evaluating the quality of AI-generated videos,\nplease watch the following frames of a given video and see the text prompt for generating the video,\nthen give scores from 5 different dimensions:\n(1) visual quality: the quality of the video in terms of clearness, resolution, brightness, and color\n(2) temporal consistency, both the consistency of objects or humans and the smoothness of motion or movements\n(3) dynamic degree, the degree of dynamic changes\n(4) text-to-video alignment, the alignment between the text prompt and the video content\n(5) factual consistency, the consistency of the video content with the common-sense and factual knowledge\n\nfor each dimension, output a float number from 1.0 to 4.0,\nthe higher the number is, the better the video performs in that sub-score, \nthe lowest 1.0 means Bad, the highest 4.0 means Perfect/Real (the video is like a real video)\nHere is an output example:\nvisual quality: 3.2\ntemporal consistency: 2.7\ndynamic degree: 4.0\ntext-to-video alignment: 2.3\nfactual consistency: 1.8\n\nFor this video, the text prompt is \"fly through tour of a museum with many paintings and sculptures and beautiful works of art in all styles\",\nall the frames of video are as follows:\n\n<image> <image> <image> <image> <image> <image> <image> <image> " }, { "from": "gpt", "value": "visual quality: 2\n temporal consistency: 2\n dynamic degree: 3\n text-to-video alignment: 0\n factual consistency: 3\n" } ]

[ "2000414_00.jpg", "2000414_01.jpg", "2000414_02.jpg", "2000414_03.jpg", "2000414_04.jpg", "2000414_05.jpg", "2000414_06.jpg", "2000414_07.jpg", "2000414_08.jpg", "2000414_09.jpg", "2000414_10.jpg", "2000414_11.jpg", "2000414_12.jpg", "2000414_13.jpg", "2000414_14.jpg", "2000414_15.jpg" ]

[ { "from": "human", "value": "\nSuppose you are an expert in judging and evaluating the quality of AI-generated videos,\nplease watch the following frames of a given video and see the text prompt for generating the video,\nthen give scores from 5 different dimensions:\n(1) visual quality: the quality of the video in terms of clearness, resolution, brightness, and color\n(2) temporal consistency, both the consistency of objects or humans and the smoothness of motion or movements\n(3) dynamic degree, the degree of dynamic changes\n(4) text-to-video alignment, the alignment between the text prompt and the video content\n(5) factual consistency, the consistency of the video content with the common-sense and factual knowledge\n\nfor each dimension, output a float number from 1.0 to 4.0,\nthe higher the number is, the better the video performs in that sub-score, \nthe lowest 1.0 means Bad, the highest 4.0 means Perfect/Real (the video is like a real video)\nHere is an output example:\nvisual quality: 3.2\ntemporal consistency: 2.7\ndynamic degree: 4.0\ntext-to-video alignment: 2.3\nfactual consistency: 1.8\n\nFor this video, the text prompt is \"a video of forest foggy environment and red blue sky and cloudy and sunset\",\nall the frames of video are as follows:\n\n<image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> " }, { "from": "gpt", "value": "visual quality: 3\n temporal consistency: 3\n dynamic degree: 3\n text-to-video alignment: 0\n factual consistency: 3\n" } ]

[ "5000772_00.jpg", "5000772_01.jpg", "5000772_02.jpg", "5000772_03.jpg", "5000772_04.jpg", "5000772_05.jpg", "5000772_06.jpg", "5000772_07.jpg", "5000772_08.jpg", "5000772_09.jpg", "5000772_10.jpg", "5000772_11.jpg", "5000772_12.jpg", "5000772_13.jpg", "5000772_14.jpg", "5000772_15.jpg" ]

[ { "from": "human", "value": "\nSuppose you are an expert in judging and evaluating the quality of AI-generated videos,\nplease watch the following frames of a given video and see the text prompt for generating the video,\nthen give scores from 5 different dimensions:\n(1) visual quality: the quality of the video in terms of clearness, resolution, brightness, and color\n(2) temporal consistency, both the consistency of objects or humans and the smoothness of motion or movements\n(3) dynamic degree, the degree of dynamic changes\n(4) text-to-video alignment, the alignment between the text prompt and the video content\n(5) factual consistency, the consistency of the video content with the common-sense and factual knowledge\n\nfor each dimension, output a float number from 1.0 to 4.0,\nthe higher the number is, the better the video performs in that sub-score, \nthe lowest 1.0 means Bad, the highest 4.0 means Perfect/Real (the video is like a real video)\nHere is an output example:\nvisual quality: 3.2\ntemporal consistency: 2.7\ndynamic degree: 4.0\ntext-to-video alignment: 2.3\nfactual consistency: 1.8\n\nFor this video, the text prompt is \"Godzilla on a slide, screaming of fun\",\nall the frames of video are as follows:\n\n<image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> " }, { "from": "gpt", "value": "visual quality: 3\n temporal consistency: 3\n dynamic degree: 1\n text-to-video alignment: 0\n factual consistency: 3\n" } ]

[ "0001694_00.jpg", "0001694_01.jpg", "0001694_02.jpg", "0001694_03.jpg", "0001694_04.jpg", "0001694_05.jpg", "0001694_06.jpg", "0001694_07.jpg", "0001694_08.jpg", "0001694_09.jpg", "0001694_10.jpg", "0001694_11.jpg", "0001694_12.jpg", "0001694_13.jpg", "0001694_14.jpg", "0001694_15.jpg", "0001694_16.jpg", "0001694_17.jpg", "0001694_18.jpg", "0001694_19.jpg", "0001694_20.jpg", "0001694_21.jpg", "0001694_22.jpg", "0001694_23.jpg" ]

[ { "from": "human", "value": "\nSuppose you are an expert in judging and evaluating the quality of AI-generated videos,\nplease watch the following frames of a given video and see the text prompt for generating the video,\nthen give scores from 5 different dimensions:\n(1) visual quality: the quality of the video in terms of clearness, resolution, brightness, and color\n(2) temporal consistency, both the consistency of objects or humans and the smoothness of motion or movements\n(3) dynamic degree, the degree of dynamic changes\n(4) text-to-video alignment, the alignment between the text prompt and the video content\n(5) factual consistency, the consistency of the video content with the common-sense and factual knowledge\n\nfor each dimension, output a float number from 1.0 to 4.0,\nthe higher the number is, the better the video performs in that sub-score, \nthe lowest 1.0 means Bad, the highest 4.0 means Perfect/Real (the video is like a real video)\nHere is an output example:\nvisual quality: 3.2\ntemporal consistency: 2.7\ndynamic degree: 4.0\ntext-to-video alignment: 2.3\nfactual consistency: 1.8\n\nFor this video, the text prompt is \"girl sitting at desk reading bible in cozy room, fireplace, Christmas tree\",\nall the frames of video are as follows:\n\n<image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> " }, { "from": "gpt", "value": "visual quality: 3\n temporal consistency: 3\n dynamic degree: 1\n text-to-video alignment: 0\n factual consistency: 3\n" } ]

[ "2000711_00.jpg", "2000711_01.jpg", "2000711_02.jpg", "2000711_03.jpg", "2000711_04.jpg", "2000711_05.jpg", "2000711_06.jpg", "2000711_07.jpg", "2000711_08.jpg", "2000711_09.jpg", "2000711_10.jpg", "2000711_11.jpg", "2000711_12.jpg", "2000711_13.jpg", "2000711_14.jpg", "2000711_15.jpg" ]

[ { "from": "human", "value": "\nSuppose you are an expert in judging and evaluating the quality of AI-generated videos,\nplease watch the following frames of a given video and see the text prompt for generating the video,\nthen give scores from 5 different dimensions:\n(1) visual quality: the quality of the video in terms of clearness, resolution, brightness, and color\n(2) temporal consistency, both the consistency of objects or humans and the smoothness of motion or movements\n(3) dynamic degree, the degree of dynamic changes\n(4) text-to-video alignment, the alignment between the text prompt and the video content\n(5) factual consistency, the consistency of the video content with the common-sense and factual knowledge\n\nfor each dimension, output a float number from 1.0 to 4.0,\nthe higher the number is, the better the video performs in that sub-score, \nthe lowest 1.0 means Bad, the highest 4.0 means Perfect/Real (the video is like a real video)\nHere is an output example:\nvisual quality: 3.2\ntemporal consistency: 2.7\ndynamic degree: 4.0\ntext-to-video alignment: 2.3\nfactual consistency: 1.8\n\nFor this video, the text prompt is \"vampire standing in love standing front of castle cover art\",\nall the frames of video are as follows:\n\n<image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> " }, { "from": "gpt", "value": "visual quality: 3\n temporal consistency: 3\n dynamic degree: 3\n text-to-video alignment: 0\n factual consistency: 2\n" } ]

[ "8003635_00.jpg", "8003635_01.jpg", "8003635_02.jpg", "8003635_03.jpg", "8003635_04.jpg", "8003635_05.jpg", "8003635_06.jpg", "8003635_07.jpg", "8003635_08.jpg", "8003635_09.jpg", "8003635_10.jpg", "8003635_11.jpg", "8003635_12.jpg", "8003635_13.jpg", "8003635_14.jpg", "8003635_15.jpg", "8003635_16.jpg", "8003635_17.jpg", "8003635_18.jpg", "8003635_19.jpg", "8003635_20.jpg", "8003635_21.jpg", "8003635_22.jpg", "8003635_23.jpg" ]

[ { "from": "human", "value": "\nSuppose you are an expert in judging and evaluating the quality of AI-generated videos,\nplease watch the following frames of a given video and see the text prompt for generating the video,\nthen give scores from 5 different dimensions:\n(1) visual quality: the quality of the video in terms of clearness, resolution, brightness, and color\n(2) temporal consistency, both the consistency of objects or humans and the smoothness of motion or movements\n(3) dynamic degree, the degree of dynamic changes\n(4) text-to-video alignment, the alignment between the text prompt and the video content\n(5) factual consistency, the consistency of the video content with the common-sense and factual knowledge\n\nfor each dimension, output a float number from 1.0 to 4.0,\nthe higher the number is, the better the video performs in that sub-score, \nthe lowest 1.0 means Bad, the highest 4.0 means Perfect/Real (the video is like a real video)\nHere is an output example:\nvisual quality: 3.2\ntemporal consistency: 2.7\ndynamic degree: 4.0\ntext-to-video alignment: 2.3\nfactual consistency: 1.8\n\nFor this video, the text prompt is \"Clown fish swimming through the coral reef\",\nall the frames of video are as follows:\n\n<image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> " }, { "from": "gpt", "value": "visual quality: 1\n temporal consistency: 3\n dynamic degree: 3\n text-to-video alignment: 0\n factual consistency: 3\n" } ]

[ "3006467_00.jpg", "3006467_01.jpg", "3006467_02.jpg", "3006467_03.jpg", "3006467_04.jpg", "3006467_05.jpg", "3006467_06.jpg", "3006467_07.jpg", "3006467_08.jpg", "3006467_09.jpg", "3006467_10.jpg", "3006467_11.jpg", "3006467_12.jpg", "3006467_13.jpg", "3006467_14.jpg", "3006467_15.jpg" ]

[ { "from": "human", "value": "\nSuppose you are an expert in judging and evaluating the quality of AI-generated videos,\nplease watch the following frames of a given video and see the text prompt for generating the video,\nthen give scores from 5 different dimensions:\n(1) visual quality: the quality of the video in terms of clearness, resolution, brightness, and color\n(2) temporal consistency, both the consistency of objects or humans and the smoothness of motion or movements\n(3) dynamic degree, the degree of dynamic changes\n(4) text-to-video alignment, the alignment between the text prompt and the video content\n(5) factual consistency, the consistency of the video content with the common-sense and factual knowledge\n\nfor each dimension, output a float number from 1.0 to 4.0,\nthe higher the number is, the better the video performs in that sub-score, \nthe lowest 1.0 means Bad, the highest 4.0 means Perfect/Real (the video is like a real video)\nHere is an output example:\nvisual quality: 3.2\ntemporal consistency: 2.7\ndynamic degree: 4.0\ntext-to-video alignment: 2.3\nfactual consistency: 1.8\n\nFor this video, the text prompt is \"in the style of a graphical chart. There is an jagged arrow rising on the chart, a consumer animation is pushing a shopping cart up the arrow. The consumers features are indistinct. The overall look is two dimensional\",\nall the frames of video are as follows:\n\n<image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> " }, { "from": "gpt", "value": "visual quality: 1\n temporal consistency: 1\n dynamic degree: 3\n text-to-video alignment: 0\n factual consistency: 2\n" } ]

[ "8002036_00.jpg", "8002036_01.jpg", "8002036_02.jpg", "8002036_03.jpg", "8002036_04.jpg", "8002036_05.jpg", "8002036_06.jpg", "8002036_07.jpg", "8002036_08.jpg", "8002036_09.jpg", "8002036_10.jpg", "8002036_11.jpg", "8002036_12.jpg", "8002036_13.jpg", "8002036_14.jpg", "8002036_15.jpg", "8002036_16.jpg", "8002036_17.jpg", "8002036_18.jpg", "8002036_19.jpg", "8002036_20.jpg", "8002036_21.jpg", "8002036_22.jpg", "8002036_23.jpg" ]

[ { "from": "human", "value": "\nSuppose you are an expert in judging and evaluating the quality of AI-generated videos,\nplease watch the following frames of a given video and see the text prompt for generating the video,\nthen give scores from 5 different dimensions:\n(1) visual quality: the quality of the video in terms of clearness, resolution, brightness, and color\n(2) temporal consistency, both the consistency of objects or humans and the smoothness of motion or movements\n(3) dynamic degree, the degree of dynamic changes\n(4) text-to-video alignment, the alignment between the text prompt and the video content\n(5) factual consistency, the consistency of the video content with the common-sense and factual knowledge\n\nfor each dimension, output a float number from 1.0 to 4.0,\nthe higher the number is, the better the video performs in that sub-score, \nthe lowest 1.0 means Bad, the highest 4.0 means Perfect/Real (the video is like a real video)\nHere is an output example:\nvisual quality: 3.2\ntemporal consistency: 2.7\ndynamic degree: 4.0\ntext-to-video alignment: 2.3\nfactual consistency: 1.8\n\nFor this video, the text prompt is \"henry cavill in a superman suit flying in Speed of light in new york city in man of steel movie scene\",\nall the frames of video are as follows:\n\n<image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> " }, { "from": "gpt", "value": "visual quality: 1\n temporal consistency: 2\n dynamic degree: 3\n text-to-video alignment: 0\n factual consistency: 1\n" } ]

[ "3000149_00.jpg", "3000149_01.jpg", "3000149_02.jpg", "3000149_03.jpg", "3000149_04.jpg", "3000149_05.jpg", "3000149_06.jpg", "3000149_07.jpg", "3000149_08.jpg", "3000149_09.jpg", "3000149_10.jpg", "3000149_11.jpg", "3000149_12.jpg", "3000149_13.jpg", "3000149_14.jpg", "3000149_15.jpg" ]

[ { "from": "human", "value": "\nSuppose you are an expert in judging and evaluating the quality of AI-generated videos,\nplease watch the following frames of a given video and see the text prompt for generating the video,\nthen give scores from 5 different dimensions:\n(1) visual quality: the quality of the video in terms of clearness, resolution, brightness, and color\n(2) temporal consistency, both the consistency of objects or humans and the smoothness of motion or movements\n(3) dynamic degree, the degree of dynamic changes\n(4) text-to-video alignment, the alignment between the text prompt and the video content\n(5) factual consistency, the consistency of the video content with the common-sense and factual knowledge\n\nfor each dimension, output a float number from 1.0 to 4.0,\nthe higher the number is, the better the video performs in that sub-score, \nthe lowest 1.0 means Bad, the highest 4.0 means Perfect/Real (the video is like a real video)\nHere is an output example:\nvisual quality: 3.2\ntemporal consistency: 2.7\ndynamic degree: 4.0\ntext-to-video alignment: 2.3\nfactual consistency: 1.8\n\nFor this video, the text prompt is \"A cat running to the Animals lining up\",\nall the frames of video are as follows:\n\n<image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> " }, { "from": "gpt", "value": "visual quality: 3\n temporal consistency: 3\n dynamic degree: 1\n text-to-video alignment: 0\n factual consistency: 3\n" } ]

[ "3000303_00.jpg", "3000303_01.jpg", "3000303_02.jpg", "3000303_03.jpg", "3000303_04.jpg", "3000303_05.jpg", "3000303_06.jpg", "3000303_07.jpg", "3000303_08.jpg", "3000303_09.jpg", "3000303_10.jpg", "3000303_11.jpg", "3000303_12.jpg", "3000303_13.jpg", "3000303_14.jpg", "3000303_15.jpg" ]

[ { "from": "human", "value": "\nSuppose you are an expert in judging and evaluating the quality of AI-generated videos,\nplease watch the following frames of a given video and see the text prompt for generating the video,\nthen give scores from 5 different dimensions:\n(1) visual quality: the quality of the video in terms of clearness, resolution, brightness, and color\n(2) temporal consistency, both the consistency of objects or humans and the smoothness of motion or movements\n(3) dynamic degree, the degree of dynamic changes\n(4) text-to-video alignment, the alignment between the text prompt and the video content\n(5) factual consistency, the consistency of the video content with the common-sense and factual knowledge\n\nfor each dimension, output a float number from 1.0 to 4.0,\nthe higher the number is, the better the video performs in that sub-score, \nthe lowest 1.0 means Bad, the highest 4.0 means Perfect/Real (the video is like a real video)\nHere is an output example:\nvisual quality: 3.2\ntemporal consistency: 2.7\ndynamic degree: 4.0\ntext-to-video alignment: 2.3\nfactual consistency: 1.8\n\nFor this video, the text prompt is \"In the serene waters of North Wales, a family embarks on a summer holiday, unaware of the tempest about to shatter their lives.\",\nall the frames of video are as follows:\n\n<image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> " }, { "from": "gpt", "value": "visual quality: 2\n temporal consistency: 2\n dynamic degree: 3\n text-to-video alignment: 0\n factual consistency: 2\n" } ]

[ "1002453_00.jpg", "1002453_01.jpg", "1002453_02.jpg", "1002453_03.jpg", "1002453_04.jpg", "1002453_05.jpg", "1002453_06.jpg", "1002453_07.jpg", "1002453_08.jpg", "1002453_09.jpg", "1002453_10.jpg", "1002453_11.jpg", "1002453_12.jpg", "1002453_13.jpg", "1002453_14.jpg", "1002453_15.jpg" ]

[ { "from": "human", "value": "\nSuppose you are an expert in judging and evaluating the quality of AI-generated videos,\nplease watch the following frames of a given video and see the text prompt for generating the video,\nthen give scores from 5 different dimensions:\n(1) visual quality: the quality of the video in terms of clearness, resolution, brightness, and color\n(2) temporal consistency, both the consistency of objects or humans and the smoothness of motion or movements\n(3) dynamic degree, the degree of dynamic changes\n(4) text-to-video alignment, the alignment between the text prompt and the video content\n(5) factual consistency, the consistency of the video content with the common-sense and factual knowledge\n\nfor each dimension, output a float number from 1.0 to 4.0,\nthe higher the number is, the better the video performs in that sub-score, \nthe lowest 1.0 means Bad, the highest 4.0 means Perfect/Real (the video is like a real video)\nHere is an output example:\nvisual quality: 3.2\ntemporal consistency: 2.7\ndynamic degree: 4.0\ntext-to-video alignment: 2.3\nfactual consistency: 1.8\n\nFor this video, the text prompt is \"Visualize the commitment of the early Church to public prayer and fasting, leading to the formation of new congregations (Acts 14:23).\",\nall the frames of video are as follows:\n\n<image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> " }, { "from": "gpt", "value": "visual quality: 1\n temporal consistency: 1\n dynamic degree: 3\n text-to-video alignment: 0\n factual consistency: 1\n" } ]

[ "2001422_00.jpg", "2001422_01.jpg", "2001422_02.jpg", "2001422_03.jpg", "2001422_04.jpg", "2001422_05.jpg", "2001422_06.jpg", "2001422_07.jpg", "2001422_08.jpg", "2001422_09.jpg", "2001422_10.jpg", "2001422_11.jpg", "2001422_12.jpg", "2001422_13.jpg", "2001422_14.jpg", "2001422_15.jpg" ]

[ { "from": "human", "value": "\nSuppose you are an expert in judging and evaluating the quality of AI-generated videos,\nplease watch the following frames of a given video and see the text prompt for generating the video,\nthen give scores from 5 different dimensions:\n(1) visual quality: the quality of the video in terms of clearness, resolution, brightness, and color\n(2) temporal consistency, both the consistency of objects or humans and the smoothness of motion or movements\n(3) dynamic degree, the degree of dynamic changes\n(4) text-to-video alignment, the alignment between the text prompt and the video content\n(5) factual consistency, the consistency of the video content with the common-sense and factual knowledge\n\nfor each dimension, output a float number from 1.0 to 4.0,\nthe higher the number is, the better the video performs in that sub-score, \nthe lowest 1.0 means Bad, the highest 4.0 means Perfect/Real (the video is like a real video)\nHere is an output example:\nvisual quality: 3.2\ntemporal consistency: 2.7\ndynamic degree: 4.0\ntext-to-video alignment: 2.3\nfactual consistency: 1.8\n\nFor this video, the text prompt is \"a man reading book in the space\",\nall the frames of video are as follows:\n\n<image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> " }, { "from": "gpt", "value": "visual quality: 3\n temporal consistency: 3\n dynamic degree: 3\n text-to-video alignment: 0\n factual consistency: 2\n" } ]

[ "5000346_00.jpg", "5000346_01.jpg", "5000346_02.jpg", "5000346_03.jpg", "5000346_04.jpg", "5000346_05.jpg", "5000346_06.jpg", "5000346_07.jpg", "5000346_08.jpg", "5000346_09.jpg", "5000346_10.jpg", "5000346_11.jpg", "5000346_12.jpg", "5000346_13.jpg", "5000346_14.jpg", "5000346_15.jpg" ]

[ { "from": "human", "value": "\nSuppose you are an expert in judging and evaluating the quality of AI-generated videos,\nplease watch the following frames of a given video and see the text prompt for generating the video,\nthen give scores from 5 different dimensions:\n(1) visual quality: the quality of the video in terms of clearness, resolution, brightness, and color\n(2) temporal consistency, both the consistency of objects or humans and the smoothness of motion or movements\n(3) dynamic degree, the degree of dynamic changes\n(4) text-to-video alignment, the alignment between the text prompt and the video content\n(5) factual consistency, the consistency of the video content with the common-sense and factual knowledge\n\nfor each dimension, output a float number from 1.0 to 4.0,\nthe higher the number is, the better the video performs in that sub-score, \nthe lowest 1.0 means Bad, the highest 4.0 means Perfect/Real (the video is like a real video)\nHere is an output example:\nvisual quality: 3.2\ntemporal consistency: 2.7\ndynamic degree: 4.0\ntext-to-video alignment: 2.3\nfactual consistency: 1.8\n\nFor this video, the text prompt is \"make the clouds above into a swirling tornado of colors and have the balloons drifting about in its wind, 4k, realism\",\nall the frames of video are as follows:\n\n<image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> " }, { "from": "gpt", "value": "visual quality: 2\n temporal consistency: 2\n dynamic degree: 3\n text-to-video alignment: 0\n factual consistency: 1\n" } ]

[ "4000610_00.jpg", "4000610_01.jpg", "4000610_02.jpg", "4000610_03.jpg", "4000610_04.jpg", "4000610_05.jpg", "4000610_06.jpg", "4000610_07.jpg", "4000610_08.jpg", "4000610_09.jpg", "4000610_10.jpg", "4000610_11.jpg", "4000610_12.jpg", "4000610_13.jpg", "4000610_14.jpg", "4000610_15.jpg" ]

[ { "from": "human", "value": "\nSuppose you are an expert in judging and evaluating the quality of AI-generated videos,\nplease watch the following frames of a given video and see the text prompt for generating the video,\nthen give scores from 5 different dimensions:\n(1) visual quality: the quality of the video in terms of clearness, resolution, brightness, and color\n(2) temporal consistency, both the consistency of objects or humans and the smoothness of motion or movements\n(3) dynamic degree, the degree of dynamic changes\n(4) text-to-video alignment, the alignment between the text prompt and the video content\n(5) factual consistency, the consistency of the video content with the common-sense and factual knowledge\n\nfor each dimension, output a float number from 1.0 to 4.0,\nthe higher the number is, the better the video performs in that sub-score, \nthe lowest 1.0 means Bad, the highest 4.0 means Perfect/Real (the video is like a real video)\nHere is an output example:\nvisual quality: 3.2\ntemporal consistency: 2.7\ndynamic degree: 4.0\ntext-to-video alignment: 2.3\nfactual consistency: 1.8\n\nFor this video, the text prompt is \"he scene shifts to the home of Xiao Meng and Xiao Yong, two panda siblings awakening in the sunlight to start their day.\",\nall the frames of video are as follows:\n\n<image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> " }, { "from": "gpt", "value": "visual quality: 3\n temporal consistency: 3\n dynamic degree: 3\n text-to-video alignment: 0\n factual consistency: 1\n" } ]

[ "4004630_00.jpg", "4004630_01.jpg", "4004630_02.jpg", "4004630_03.jpg", "4004630_04.jpg", "4004630_05.jpg", "4004630_06.jpg", "4004630_07.jpg", "4004630_08.jpg", "4004630_09.jpg", "4004630_10.jpg", "4004630_11.jpg", "4004630_12.jpg", "4004630_13.jpg", "4004630_14.jpg", "4004630_15.jpg" ]

[ { "from": "human", "value": "\nSuppose you are an expert in judging and evaluating the quality of AI-generated videos,\nplease watch the following frames of a given video and see the text prompt for generating the video,\nthen give scores from 5 different dimensions:\n(1) visual quality: the quality of the video in terms of clearness, resolution, brightness, and color\n(2) temporal consistency, both the consistency of objects or humans and the smoothness of motion or movements\n(3) dynamic degree, the degree of dynamic changes\n(4) text-to-video alignment, the alignment between the text prompt and the video content\n(5) factual consistency, the consistency of the video content with the common-sense and factual knowledge\n\nfor each dimension, output a float number from 1.0 to 4.0,\nthe higher the number is, the better the video performs in that sub-score, \nthe lowest 1.0 means Bad, the highest 4.0 means Perfect/Real (the video is like a real video)\nHere is an output example:\nvisual quality: 3.2\ntemporal consistency: 2.7\ndynamic degree: 4.0\ntext-to-video alignment: 2.3\nfactual consistency: 1.8\n\nFor this video, the text prompt is \"Person or business gasping for breath 3d animiert cinema\",\nall the frames of video are as follows:\n\n<image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> " }, { "from": "gpt", "value": "visual quality: 3\n temporal consistency: 2\n dynamic degree: 3\n text-to-video alignment: 0\n factual consistency: 2\n" } ]

[ "8003638_00.jpg", "8003638_01.jpg", "8003638_02.jpg", "8003638_03.jpg", "8003638_04.jpg", "8003638_05.jpg", "8003638_06.jpg", "8003638_07.jpg", "8003638_08.jpg", "8003638_09.jpg", "8003638_10.jpg", "8003638_11.jpg", "8003638_12.jpg", "8003638_13.jpg", "8003638_14.jpg", "8003638_15.jpg", "8003638_16.jpg", "8003638_17.jpg", "8003638_18.jpg", "8003638_19.jpg", "8003638_20.jpg", "8003638_21.jpg", "8003638_22.jpg", "8003638_23.jpg" ]

[ { "from": "human", "value": "\nSuppose you are an expert in judging and evaluating the quality of AI-generated videos,\nplease watch the following frames of a given video and see the text prompt for generating the video,\nthen give scores from 5 different dimensions:\n(1) visual quality: the quality of the video in terms of clearness, resolution, brightness, and color\n(2) temporal consistency, both the consistency of objects or humans and the smoothness of motion or movements\n(3) dynamic degree, the degree of dynamic changes\n(4) text-to-video alignment, the alignment between the text prompt and the video content\n(5) factual consistency, the consistency of the video content with the common-sense and factual knowledge\n\nfor each dimension, output a float number from 1.0 to 4.0,\nthe higher the number is, the better the video performs in that sub-score, \nthe lowest 1.0 means Bad, the highest 4.0 means Perfect/Real (the video is like a real video)\nHere is an output example:\nvisual quality: 3.2\ntemporal consistency: 2.7\ndynamic degree: 4.0\ntext-to-video alignment: 2.3\nfactual consistency: 1.8\n\nFor this video, the text prompt is \"completely base your choice of which one to visit today on the dish that most entices your taste buds, 1080P, high quality, comic\",\nall the frames of video are as follows:\n\n<image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> " }, { "from": "gpt", "value": "visual quality: 1\n temporal consistency: 1\n dynamic degree: 3\n text-to-video alignment: 0\n factual consistency: 1\n" } ]

[ "9000390_00.jpg", "9000390_01.jpg", "9000390_02.jpg", "9000390_03.jpg", "9000390_04.jpg", "9000390_05.jpg", "9000390_06.jpg", "9000390_07.jpg", "9000390_08.jpg", "9000390_09.jpg", "9000390_10.jpg", "9000390_11.jpg", "9000390_12.jpg", "9000390_13.jpg", "9000390_14.jpg", "9000390_15.jpg", "9000390_16.jpg", "9000390_17.jpg", "9000390_18.jpg", "9000390_19.jpg", "9000390_20.jpg", "9000390_21.jpg", "9000390_22.jpg", "9000390_23.jpg" ]

[ { "from": "human", "value": "\nSuppose you are an expert in judging and evaluating the quality of AI-generated videos,\nplease watch the following frames of a given video and see the text prompt for generating the video,\nthen give scores from 5 different dimensions:\n(1) visual quality: the quality of the video in terms of clearness, resolution, brightness, and color\n(2) temporal consistency, both the consistency of objects or humans and the smoothness of motion or movements\n(3) dynamic degree, the degree of dynamic changes\n(4) text-to-video alignment, the alignment between the text prompt and the video content\n(5) factual consistency, the consistency of the video content with the common-sense and factual knowledge\n\nfor each dimension, output a float number from 1.0 to 4.0,\nthe higher the number is, the better the video performs in that sub-score, \nthe lowest 1.0 means Bad, the highest 4.0 means Perfect/Real (the video is like a real video)\nHere is an output example:\nvisual quality: 3.2\ntemporal consistency: 2.7\ndynamic degree: 4.0\ntext-to-video alignment: 2.3\nfactual consistency: 1.8\n\nFor this video, the text prompt is \"women in pink niqab running up a mountain made of gold coins\",\nall the frames of video are as follows:\n\n<image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> " }, { "from": "gpt", "value": "visual quality: 3\n temporal consistency: 3\n dynamic degree: 3\n text-to-video alignment: 0\n factual consistency: 3\n" } ]

[ "8002923_00.jpg", "8002923_01.jpg", "8002923_02.jpg", "8002923_03.jpg", "8002923_04.jpg", "8002923_05.jpg", "8002923_06.jpg", "8002923_07.jpg", "8002923_08.jpg", "8002923_09.jpg", "8002923_10.jpg", "8002923_11.jpg", "8002923_12.jpg", "8002923_13.jpg", "8002923_14.jpg", "8002923_15.jpg", "8002923_16.jpg", "8002923_17.jpg", "8002923_18.jpg", "8002923_19.jpg", "8002923_20.jpg", "8002923_21.jpg", "8002923_22.jpg", "8002923_23.jpg" ]

[ { "from": "human", "value": "\nSuppose you are an expert in judging and evaluating the quality of AI-generated videos,\nplease watch the following frames of a given video and see the text prompt for generating the video,\nthen give scores from 5 different dimensions:\n(1) visual quality: the quality of the video in terms of clearness, resolution, brightness, and color\n(2) temporal consistency, both the consistency of objects or humans and the smoothness of motion or movements\n(3) dynamic degree, the degree of dynamic changes\n(4) text-to-video alignment, the alignment between the text prompt and the video content\n(5) factual consistency, the consistency of the video content with the common-sense and factual knowledge\n\nfor each dimension, output a float number from 1.0 to 4.0,\nthe higher the number is, the better the video performs in that sub-score, \nthe lowest 1.0 means Bad, the highest 4.0 means Perfect/Real (the video is like a real video)\nHere is an output example:\nvisual quality: 3.2\ntemporal consistency: 2.7\ndynamic degree: 4.0\ntext-to-video alignment: 2.3\nfactual consistency: 1.8\n\nFor this video, the text prompt is \"Person or business gasping for breath 3d animiert cinema\",\nall the frames of video are as follows:\n\n<image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> " }, { "from": "gpt", "value": "visual quality: 1\n temporal consistency: 1\n dynamic degree: 3\n text-to-video alignment: 0\n factual consistency: 1\n" } ]

[ "6000906_00.jpg", "6000906_01.jpg", "6000906_02.jpg", "6000906_03.jpg", "6000906_04.jpg", "6000906_05.jpg", "6000906_06.jpg", "6000906_07.jpg", "6000906_08.jpg", "6000906_09.jpg", "6000906_10.jpg", "6000906_11.jpg", "6000906_12.jpg", "6000906_13.jpg", "6000906_14.jpg", "6000906_15.jpg" ]

[ { "from": "human", "value": "\nSuppose you are an expert in judging and evaluating the quality of AI-generated videos,\nplease watch the following frames of a given video and see the text prompt for generating the video,\nthen give scores from 5 different dimensions:\n(1) visual quality: the quality of the video in terms of clearness, resolution, brightness, and color\n(2) temporal consistency, both the consistency of objects or humans and the smoothness of motion or movements\n(3) dynamic degree, the degree of dynamic changes\n(4) text-to-video alignment, the alignment between the text prompt and the video content\n(5) factual consistency, the consistency of the video content with the common-sense and factual knowledge\n\nfor each dimension, output a float number from 1.0 to 4.0,\nthe higher the number is, the better the video performs in that sub-score, \nthe lowest 1.0 means Bad, the highest 4.0 means Perfect/Real (the video is like a real video)\nHere is an output example:\nvisual quality: 3.2\ntemporal consistency: 2.7\ndynamic degree: 4.0\ntext-to-video alignment: 2.3\nfactual consistency: 1.8\n\nFor this video, the text prompt is \"A car driving on a highway at night. 1984 movie directed by Steven Spielberg\",\nall the frames of video are as follows:\n\n<image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> " }, { "from": "gpt", "value": "visual quality: 1\n temporal consistency: 1\n dynamic degree: 3\n text-to-video alignment: 0\n factual consistency: 1\n" } ]

[ "1001511_00.jpg", "1001511_01.jpg", "1001511_02.jpg", "1001511_03.jpg", "1001511_04.jpg", "1001511_05.jpg", "1001511_06.jpg", "1001511_07.jpg", "1001511_08.jpg", "1001511_09.jpg", "1001511_10.jpg", "1001511_11.jpg", "1001511_12.jpg", "1001511_13.jpg", "1001511_14.jpg", "1001511_15.jpg" ]

[ { "from": "human", "value": "\nSuppose you are an expert in judging and evaluating the quality of AI-generated videos,\nplease watch the following frames of a given video and see the text prompt for generating the video,\nthen give scores from 5 different dimensions:\n(1) visual quality: the quality of the video in terms of clearness, resolution, brightness, and color\n(2) temporal consistency, both the consistency of objects or humans and the smoothness of motion or movements\n(3) dynamic degree, the degree of dynamic changes\n(4) text-to-video alignment, the alignment between the text prompt and the video content\n(5) factual consistency, the consistency of the video content with the common-sense and factual knowledge\n\nfor each dimension, output a float number from 1.0 to 4.0,\nthe higher the number is, the better the video performs in that sub-score, \nthe lowest 1.0 means Bad, the highest 4.0 means Perfect/Real (the video is like a real video)\nHere is an output example:\nvisual quality: 3.2\ntemporal consistency: 2.7\ndynamic degree: 4.0\ntext-to-video alignment: 2.3\nfactual consistency: 1.8\n\nFor this video, the text prompt is \"Minotaur holds an ax in his hand\",\nall the frames of video are as follows:\n\n<image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> " }, { "from": "gpt", "value": "visual quality: 1\n temporal consistency: 1\n dynamic degree: 3\n text-to-video alignment: 0\n factual consistency: 1\n" } ]

[ "0005318_00.jpg", "0005318_01.jpg", "0005318_02.jpg", "0005318_03.jpg", "0005318_04.jpg", "0005318_05.jpg", "0005318_06.jpg", "0005318_07.jpg", "0005318_08.jpg", "0005318_09.jpg", "0005318_10.jpg", "0005318_11.jpg", "0005318_12.jpg", "0005318_13.jpg", "0005318_14.jpg", "0005318_15.jpg", "0005318_16.jpg", "0005318_17.jpg", "0005318_18.jpg", "0005318_19.jpg", "0005318_20.jpg", "0005318_21.jpg", "0005318_22.jpg", "0005318_23.jpg" ]

[ { "from": "human", "value": "\nSuppose you are an expert in judging and evaluating the quality of AI-generated videos,\nplease watch the following frames of a given video and see the text prompt for generating the video,\nthen give scores from 5 different dimensions:\n(1) visual quality: the quality of the video in terms of clearness, resolution, brightness, and color\n(2) temporal consistency, both the consistency of objects or humans and the smoothness of motion or movements\n(3) dynamic degree, the degree of dynamic changes\n(4) text-to-video alignment, the alignment between the text prompt and the video content\n(5) factual consistency, the consistency of the video content with the common-sense and factual knowledge\n\nfor each dimension, output a float number from 1.0 to 4.0,\nthe higher the number is, the better the video performs in that sub-score, \nthe lowest 1.0 means Bad, the highest 4.0 means Perfect/Real (the video is like a real video)\nHere is an output example:\nvisual quality: 3.2\ntemporal consistency: 2.7\ndynamic degree: 4.0\ntext-to-video alignment: 2.3\nfactual consistency: 1.8\n\nFor this video, the text prompt is \"create me a landscape of mars as if you were looking out of a window\",\nall the frames of video are as follows:\n\n<image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> " }, { "from": "gpt", "value": "visual quality: 2\n temporal consistency: 3\n dynamic degree: 3\n text-to-video alignment: 0\n factual consistency: 3\n" } ]

[ "2001116_00.jpg", "2001116_01.jpg", "2001116_02.jpg", "2001116_03.jpg", "2001116_04.jpg", "2001116_05.jpg", "2001116_06.jpg", "2001116_07.jpg", "2001116_08.jpg", "2001116_09.jpg", "2001116_10.jpg", "2001116_11.jpg", "2001116_12.jpg", "2001116_13.jpg", "2001116_14.jpg", "2001116_15.jpg" ]

[ { "from": "human", "value": "\nSuppose you are an expert in judging and evaluating the quality of AI-generated videos,\nplease watch the following frames of a given video and see the text prompt for generating the video,\nthen give scores from 5 different dimensions:\n(1) visual quality: the quality of the video in terms of clearness, resolution, brightness, and color\n(2) temporal consistency, both the consistency of objects or humans and the smoothness of motion or movements\n(3) dynamic degree, the degree of dynamic changes\n(4) text-to-video alignment, the alignment between the text prompt and the video content\n(5) factual consistency, the consistency of the video content with the common-sense and factual knowledge\n\nfor each dimension, output a float number from 1.0 to 4.0,\nthe higher the number is, the better the video performs in that sub-score, \nthe lowest 1.0 means Bad, the highest 4.0 means Perfect/Real (the video is like a real video)\nHere is an output example:\nvisual quality: 3.2\ntemporal consistency: 2.7\ndynamic degree: 4.0\ntext-to-video alignment: 2.3\nfactual consistency: 1.8\n\nFor this video, the text prompt is \"one cat look like a doctor,and wear white coat ,she is look forward,Background in a hospital.\",\nall the frames of video are as follows:\n\n<image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> " }, { "from": "gpt", "value": "visual quality: 3\n temporal consistency: 2\n dynamic degree: 3\n text-to-video alignment: 0\n factual consistency: 2\n" } ]

[ "4004238_00.jpg", "4004238_01.jpg", "4004238_02.jpg", "4004238_03.jpg", "4004238_04.jpg", "4004238_05.jpg", "4004238_06.jpg", "4004238_07.jpg", "4004238_08.jpg", "4004238_09.jpg", "4004238_10.jpg", "4004238_11.jpg", "4004238_12.jpg", "4004238_13.jpg", "4004238_14.jpg", "4004238_15.jpg" ]

[ { "from": "human", "value": "\nSuppose you are an expert in judging and evaluating the quality of AI-generated videos,\nplease watch the following frames of a given video and see the text prompt for generating the video,\nthen give scores from 5 different dimensions:\n(1) visual quality: the quality of the video in terms of clearness, resolution, brightness, and color\n(2) temporal consistency, both the consistency of objects or humans and the smoothness of motion or movements\n(3) dynamic degree, the degree of dynamic changes\n(4) text-to-video alignment, the alignment between the text prompt and the video content\n(5) factual consistency, the consistency of the video content with the common-sense and factual knowledge\n\nfor each dimension, output a float number from 1.0 to 4.0,\nthe higher the number is, the better the video performs in that sub-score, \nthe lowest 1.0 means Bad, the highest 4.0 means Perfect/Real (the video is like a real video)\nHere is an output example:\nvisual quality: 3.2\ntemporal consistency: 2.7\ndynamic degree: 4.0\ntext-to-video alignment: 2.3\nfactual consistency: 1.8\n\nFor this video, the text prompt is \"cozy rooms rain falling outside, to create a relaxing atmosphere for studying or unwinding.\",\nall the frames of video are as follows:\n\n<image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> " }, { "from": "gpt", "value": "visual quality: 2\n temporal consistency: 2\n dynamic degree: 3\n text-to-video alignment: 0\n factual consistency: 1\n" } ]

[ "3002950_00.jpg", "3002950_01.jpg", "3002950_02.jpg", "3002950_03.jpg", "3002950_04.jpg", "3002950_05.jpg", "3002950_06.jpg", "3002950_07.jpg", "3002950_08.jpg", "3002950_09.jpg", "3002950_10.jpg", "3002950_11.jpg", "3002950_12.jpg", "3002950_13.jpg", "3002950_14.jpg", "3002950_15.jpg" ]

[ { "from": "human", "value": "\nSuppose you are an expert in judging and evaluating the quality of AI-generated videos,\nplease watch the following frames of a given video and see the text prompt for generating the video,\nthen give scores from 5 different dimensions:\n(1) visual quality: the quality of the video in terms of clearness, resolution, brightness, and color\n(2) temporal consistency, both the consistency of objects or humans and the smoothness of motion or movements\n(3) dynamic degree, the degree of dynamic changes\n(4) text-to-video alignment, the alignment between the text prompt and the video content\n(5) factual consistency, the consistency of the video content with the common-sense and factual knowledge\n\nfor each dimension, output a float number from 1.0 to 4.0,\nthe higher the number is, the better the video performs in that sub-score, \nthe lowest 1.0 means Bad, the highest 4.0 means Perfect/Real (the video is like a real video)\nHere is an output example:\nvisual quality: 3.2\ntemporal consistency: 2.7\ndynamic degree: 4.0\ntext-to-video alignment: 2.3\nfactual consistency: 1.8\n\nFor this video, the text prompt is \"bright sunny morning Winnie the pooh With Owl leading the way, Pooh, Piglet, Eeyore, and Tigger set off towards the Great Upstream Dam. They trekked through the forest, their determination growing with every step.\",\nall the frames of video are as follows:\n\n<image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> " }, { "from": "gpt", "value": "visual quality: 3\n temporal consistency: 2\n dynamic degree: 3\n text-to-video alignment: 0\n factual consistency: 2\n" } ]

[ "9000419_00.jpg", "9000419_01.jpg", "9000419_02.jpg", "9000419_03.jpg", "9000419_04.jpg", "9000419_05.jpg", "9000419_06.jpg", "9000419_07.jpg", "9000419_08.jpg", "9000419_09.jpg", "9000419_10.jpg", "9000419_11.jpg", "9000419_12.jpg", "9000419_13.jpg", "9000419_14.jpg", "9000419_15.jpg", "9000419_16.jpg", "9000419_17.jpg", "9000419_18.jpg", "9000419_19.jpg", "9000419_20.jpg", "9000419_21.jpg", "9000419_22.jpg", "9000419_23.jpg" ]

[ { "from": "human", "value": "\nSuppose you are an expert in judging and evaluating the quality of AI-generated videos,\nplease watch the following frames of a given video and see the text prompt for generating the video,\nthen give scores from 5 different dimensions:\n(1) visual quality: the quality of the video in terms of clearness, resolution, brightness, and color\n(2) temporal consistency, both the consistency of objects or humans and the smoothness of motion or movements\n(3) dynamic degree, the degree of dynamic changes\n(4) text-to-video alignment, the alignment between the text prompt and the video content\n(5) factual consistency, the consistency of the video content with the common-sense and factual knowledge\n\nfor each dimension, output a float number from 1.0 to 4.0,\nthe higher the number is, the better the video performs in that sub-score, \nthe lowest 1.0 means Bad, the highest 4.0 means Perfect/Real (the video is like a real video)\nHere is an output example:\nvisual quality: 3.2\ntemporal consistency: 2.7\ndynamic degree: 4.0\ntext-to-video alignment: 2.3\nfactual consistency: 1.8\n\nFor this video, the text prompt is \"The menacing crocodile lurking beneath the surface of Bloody Lake.\",\nall the frames of video are as follows:\n\n<image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> " }, { "from": "gpt", "value": "visual quality: 1\n temporal consistency: 1\n dynamic degree: 3\n text-to-video alignment: 0\n factual consistency: 1\n" } ]

[ "4003597_00.jpg", "4003597_01.jpg", "4003597_02.jpg", "4003597_03.jpg", "4003597_04.jpg", "4003597_05.jpg", "4003597_06.jpg", "4003597_07.jpg", "4003597_08.jpg", "4003597_09.jpg", "4003597_10.jpg", "4003597_11.jpg", "4003597_12.jpg", "4003597_13.jpg", "4003597_14.jpg", "4003597_15.jpg" ]

[ { "from": "human", "value": "\nSuppose you are an expert in judging and evaluating the quality of AI-generated videos,\nplease watch the following frames of a given video and see the text prompt for generating the video,\nthen give scores from 5 different dimensions:\n(1) visual quality: the quality of the video in terms of clearness, resolution, brightness, and color\n(2) temporal consistency, both the consistency of objects or humans and the smoothness of motion or movements\n(3) dynamic degree, the degree of dynamic changes\n(4) text-to-video alignment, the alignment between the text prompt and the video content\n(5) factual consistency, the consistency of the video content with the common-sense and factual knowledge\n\nfor each dimension, output a float number from 1.0 to 4.0,\nthe higher the number is, the better the video performs in that sub-score, \nthe lowest 1.0 means Bad, the highest 4.0 means Perfect/Real (the video is like a real video)\nHere is an output example:\nvisual quality: 3.2\ntemporal consistency: 2.7\ndynamic degree: 4.0\ntext-to-video alignment: 2.3\nfactual consistency: 1.8\n\nFor this video, the text prompt is \"water lilies moving slowly in a serene green pond as a white butterfly slowly lands on a lilypad 16:9\",\nall the frames of video are as follows:\n\n<image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> " }, { "from": "gpt", "value": "visual quality: 3\n temporal consistency: 3\n dynamic degree: 3\n text-to-video alignment: 0\n factual consistency: 3\n" } ]

[ "5000256_00.jpg", "5000256_01.jpg", "5000256_02.jpg", "5000256_03.jpg", "5000256_04.jpg", "5000256_05.jpg", "5000256_06.jpg", "5000256_07.jpg", "5000256_08.jpg", "5000256_09.jpg", "5000256_10.jpg", "5000256_11.jpg", "5000256_12.jpg", "5000256_13.jpg", "5000256_14.jpg", "5000256_15.jpg" ]

[ { "from": "human", "value": "\nSuppose you are an expert in judging and evaluating the quality of AI-generated videos,\nplease watch the following frames of a given video and see the text prompt for generating the video,\nthen give scores from 5 different dimensions:\n(1) visual quality: the quality of the video in terms of clearness, resolution, brightness, and color\n(2) temporal consistency, both the consistency of objects or humans and the smoothness of motion or movements\n(3) dynamic degree, the degree of dynamic changes\n(4) text-to-video alignment, the alignment between the text prompt and the video content\n(5) factual consistency, the consistency of the video content with the common-sense and factual knowledge\n\nfor each dimension, output a float number from 1.0 to 4.0,\nthe higher the number is, the better the video performs in that sub-score, \nthe lowest 1.0 means Bad, the highest 4.0 means Perfect/Real (the video is like a real video)\nHere is an output example:\nvisual quality: 3.2\ntemporal consistency: 2.7\ndynamic degree: 4.0\ntext-to-video alignment: 2.3\nfactual consistency: 1.8\n\nFor this video, the text prompt is \"personn sitting in her untidy beroom\",\nall the frames of video are as follows:\n\n<image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> " }, { "from": "gpt", "value": "visual quality: 3\n temporal consistency: 1\n dynamic degree: 3\n text-to-video alignment: 0\n factual consistency: 1\n" } ]

[ "1005127_00.jpg", "1005127_01.jpg", "1005127_02.jpg", "1005127_03.jpg", "1005127_04.jpg", "1005127_05.jpg", "1005127_06.jpg", "1005127_07.jpg", "1005127_08.jpg", "1005127_09.jpg", "1005127_10.jpg", "1005127_11.jpg", "1005127_12.jpg", "1005127_13.jpg", "1005127_14.jpg", "1005127_15.jpg" ]

[ { "from": "human", "value": "\nSuppose you are an expert in judging and evaluating the quality of AI-generated videos,\nplease watch the following frames of a given video and see the text prompt for generating the video,\nthen give scores from 5 different dimensions:\n(1) visual quality: the quality of the video in terms of clearness, resolution, brightness, and color\n(2) temporal consistency, both the consistency of objects or humans and the smoothness of motion or movements\n(3) dynamic degree, the degree of dynamic changes\n(4) text-to-video alignment, the alignment between the text prompt and the video content\n(5) factual consistency, the consistency of the video content with the common-sense and factual knowledge\n\nfor each dimension, output a float number from 1.0 to 4.0,\nthe higher the number is, the better the video performs in that sub-score, \nthe lowest 1.0 means Bad, the highest 4.0 means Perfect/Real (the video is like a real video)\nHere is an output example:\nvisual quality: 3.2\ntemporal consistency: 2.7\ndynamic degree: 4.0\ntext-to-video alignment: 2.3\nfactual consistency: 1.8\n\nFor this video, the text prompt is \"The outside of a church with a rainbow flag hanging around the entire building.\",\nall the frames of video are as follows:\n\n<image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> " }, { "from": "gpt", "value": "visual quality: 1\n temporal consistency: 1\n dynamic degree: 3\n text-to-video alignment: 0\n factual consistency: 1\n" } ]

[ "2002420_00.jpg", "2002420_01.jpg", "2002420_02.jpg", "2002420_03.jpg", "2002420_04.jpg", "2002420_05.jpg", "2002420_06.jpg", "2002420_07.jpg", "2002420_08.jpg", "2002420_09.jpg", "2002420_10.jpg", "2002420_11.jpg", "2002420_12.jpg", "2002420_13.jpg", "2002420_14.jpg", "2002420_15.jpg" ]

[ { "from": "human", "value": "\nSuppose you are an expert in judging and evaluating the quality of AI-generated videos,\nplease watch the following frames of a given video and see the text prompt for generating the video,\nthen give scores from 5 different dimensions:\n(1) visual quality: the quality of the video in terms of clearness, resolution, brightness, and color\n(2) temporal consistency, both the consistency of objects or humans and the smoothness of motion or movements\n(3) dynamic degree, the degree of dynamic changes\n(4) text-to-video alignment, the alignment between the text prompt and the video content\n(5) factual consistency, the consistency of the video content with the common-sense and factual knowledge\n\nfor each dimension, output a float number from 1.0 to 4.0,\nthe higher the number is, the better the video performs in that sub-score, \nthe lowest 1.0 means Bad, the highest 4.0 means Perfect/Real (the video is like a real video)\nHere is an output example:\nvisual quality: 3.2\ntemporal consistency: 2.7\ndynamic degree: 4.0\ntext-to-video alignment: 2.3\nfactual consistency: 1.8\n\nFor this video, the text prompt is \"a japanese car, supra, with neon lights, drifting, on a mountain, night, drone shot, fast and furious scene\",\nall the frames of video are as follows:\n\n<image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> " }, { "from": "gpt", "value": "visual quality: 2\n temporal consistency: 3\n dynamic degree: 3\n text-to-video alignment: 0\n factual consistency: 1\n" } ]

[ "8000148_00.jpg", "8000148_01.jpg", "8000148_02.jpg", "8000148_03.jpg", "8000148_04.jpg", "8000148_05.jpg", "8000148_06.jpg", "8000148_07.jpg", "8000148_08.jpg", "8000148_09.jpg", "8000148_10.jpg", "8000148_11.jpg", "8000148_12.jpg", "8000148_13.jpg", "8000148_14.jpg", "8000148_15.jpg", "8000148_16.jpg", "8000148_17.jpg", "8000148_18.jpg", "8000148_19.jpg", "8000148_20.jpg", "8000148_21.jpg", "8000148_22.jpg", "8000148_23.jpg" ]

[ { "from": "human", "value": "\nSuppose you are an expert in judging and evaluating the quality of AI-generated videos,\nplease watch the following frames of a given video and see the text prompt for generating the video,\nthen give scores from 5 different dimensions:\n(1) visual quality: the quality of the video in terms of clearness, resolution, brightness, and color\n(2) temporal consistency, both the consistency of objects or humans and the smoothness of motion or movements\n(3) dynamic degree, the degree of dynamic changes\n(4) text-to-video alignment, the alignment between the text prompt and the video content\n(5) factual consistency, the consistency of the video content with the common-sense and factual knowledge\n\nfor each dimension, output a float number from 1.0 to 4.0,\nthe higher the number is, the better the video performs in that sub-score, \nthe lowest 1.0 means Bad, the highest 4.0 means Perfect/Real (the video is like a real video)\nHere is an output example:\nvisual quality: 3.2\ntemporal consistency: 2.7\ndynamic degree: 4.0\ntext-to-video alignment: 2.3\nfactual consistency: 1.8\n\nFor this video, the text prompt is \"Hip Hop dancing under a full moon on a city sidewalk\",\nall the frames of video are as follows:\n\n<image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> " }, { "from": "gpt", "value": "visual quality: 1\n temporal consistency: 1\n dynamic degree: 3\n text-to-video alignment: 0\n factual consistency: 1\n" } ]

[ "3001690_00.jpg", "3001690_01.jpg", "3001690_02.jpg", "3001690_03.jpg", "3001690_04.jpg", "3001690_05.jpg", "3001690_06.jpg", "3001690_07.jpg", "3001690_08.jpg", "3001690_09.jpg", "3001690_10.jpg", "3001690_11.jpg", "3001690_12.jpg", "3001690_13.jpg", "3001690_14.jpg", "3001690_15.jpg" ]

[ { "from": "human", "value": "\nSuppose you are an expert in judging and evaluating the quality of AI-generated videos,\nplease watch the following frames of a given video and see the text prompt for generating the video,\nthen give scores from 5 different dimensions:\n(1) visual quality: the quality of the video in terms of clearness, resolution, brightness, and color\n(2) temporal consistency, both the consistency of objects or humans and the smoothness of motion or movements\n(3) dynamic degree, the degree of dynamic changes\n(4) text-to-video alignment, the alignment between the text prompt and the video content\n(5) factual consistency, the consistency of the video content with the common-sense and factual knowledge\n\nfor each dimension, output a float number from 1.0 to 4.0,\nthe higher the number is, the better the video performs in that sub-score, \nthe lowest 1.0 means Bad, the highest 4.0 means Perfect/Real (the video is like a real video)\nHere is an output example:\nvisual quality: 3.2\ntemporal consistency: 2.7\ndynamic degree: 4.0\ntext-to-video alignment: 2.3\nfactual consistency: 1.8\n\nFor this video, the text prompt is \"a man with gun in a dark area of a city\",\nall the frames of video are as follows:\n\n<image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> " }, { "from": "gpt", "value": "visual quality: 2\n temporal consistency: 1\n dynamic degree: 3\n text-to-video alignment: 0\n factual consistency: 2\n" } ]

[ "7001909_00.jpg", "7001909_01.jpg", "7001909_02.jpg", "7001909_03.jpg", "7001909_04.jpg", "7001909_05.jpg", "7001909_06.jpg", "7001909_07.jpg" ]

[ { "from": "human", "value": "\nSuppose you are an expert in judging and evaluating the quality of AI-generated videos,\nplease watch the following frames of a given video and see the text prompt for generating the video,\nthen give scores from 5 different dimensions:\n(1) visual quality: the quality of the video in terms of clearness, resolution, brightness, and color\n(2) temporal consistency, both the consistency of objects or humans and the smoothness of motion or movements\n(3) dynamic degree, the degree of dynamic changes\n(4) text-to-video alignment, the alignment between the text prompt and the video content\n(5) factual consistency, the consistency of the video content with the common-sense and factual knowledge\n\nfor each dimension, output a float number from 1.0 to 4.0,\nthe higher the number is, the better the video performs in that sub-score, \nthe lowest 1.0 means Bad, the highest 4.0 means Perfect/Real (the video is like a real video)\nHere is an output example:\nvisual quality: 3.2\ntemporal consistency: 2.7\ndynamic degree: 4.0\ntext-to-video alignment: 2.3\nfactual consistency: 1.8\n\nFor this video, the text prompt is \"mother and daughter walking into shoe store on a bright and sunny day. very colorful\",\nall the frames of video are as follows:\n\n<image> <image> <image> <image> <image> <image> <image> <image> " }, { "from": "gpt", "value": "visual quality: 2\n temporal consistency: 2\n dynamic degree: 3\n text-to-video alignment: 0\n factual consistency: 3\n" } ]

[ "3002300_00.jpg", "3002300_01.jpg", "3002300_02.jpg", "3002300_03.jpg", "3002300_04.jpg", "3002300_05.jpg", "3002300_06.jpg", "3002300_07.jpg", "3002300_08.jpg", "3002300_09.jpg", "3002300_10.jpg", "3002300_11.jpg", "3002300_12.jpg", "3002300_13.jpg", "3002300_14.jpg", "3002300_15.jpg" ]

[ { "from": "human", "value": "\nSuppose you are an expert in judging and evaluating the quality of AI-generated videos,\nplease watch the following frames of a given video and see the text prompt for generating the video,\nthen give scores from 5 different dimensions:\n(1) visual quality: the quality of the video in terms of clearness, resolution, brightness, and color\n(2) temporal consistency, both the consistency of objects or humans and the smoothness of motion or movements\n(3) dynamic degree, the degree of dynamic changes\n(4) text-to-video alignment, the alignment between the text prompt and the video content\n(5) factual consistency, the consistency of the video content with the common-sense and factual knowledge\n\nfor each dimension, output a float number from 1.0 to 4.0,\nthe higher the number is, the better the video performs in that sub-score, \nthe lowest 1.0 means Bad, the highest 4.0 means Perfect/Real (the video is like a real video)\nHere is an output example:\nvisual quality: 3.2\ntemporal consistency: 2.7\ndynamic degree: 4.0\ntext-to-video alignment: 2.3\nfactual consistency: 1.8\n\nFor this video, the text prompt is \"night seen in man and woman image in road in car running\",\nall the frames of video are as follows:\n\n<image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> " }, { "from": "gpt", "value": "visual quality: 2\n temporal consistency: 2\n dynamic degree: 3\n text-to-video alignment: 0\n factual consistency: 2\n" } ]

[ "3000537_00.jpg", "3000537_01.jpg", "3000537_02.jpg", "3000537_03.jpg", "3000537_04.jpg", "3000537_05.jpg", "3000537_06.jpg", "3000537_07.jpg", "3000537_08.jpg", "3000537_09.jpg", "3000537_10.jpg", "3000537_11.jpg", "3000537_12.jpg", "3000537_13.jpg", "3000537_14.jpg", "3000537_15.jpg" ]

[ { "from": "human", "value": "\nSuppose you are an expert in judging and evaluating the quality of AI-generated videos,\nplease watch the following frames of a given video and see the text prompt for generating the video,\nthen give scores from 5 different dimensions:\n(1) visual quality: the quality of the video in terms of clearness, resolution, brightness, and color\n(2) temporal consistency, both the consistency of objects or humans and the smoothness of motion or movements\n(3) dynamic degree, the degree of dynamic changes\n(4) text-to-video alignment, the alignment between the text prompt and the video content\n(5) factual consistency, the consistency of the video content with the common-sense and factual knowledge\n\nfor each dimension, output a float number from 1.0 to 4.0,\nthe higher the number is, the better the video performs in that sub-score, \nthe lowest 1.0 means Bad, the highest 4.0 means Perfect/Real (the video is like a real video)\nHere is an output example:\nvisual quality: 3.2\ntemporal consistency: 2.7\ndynamic degree: 4.0\ntext-to-video alignment: 2.3\nfactual consistency: 1.8\n\nFor this video, the text prompt is \"One day the farmer fell very ill\",\nall the frames of video are as follows:\n\n<image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> " }, { "from": "gpt", "value": "visual quality: 2\n temporal consistency: 3\n dynamic degree: 3\n text-to-video alignment: 0\n factual consistency: 3\n" } ]

[ "0000806_00.jpg", "0000806_01.jpg", "0000806_02.jpg", "0000806_03.jpg", "0000806_04.jpg", "0000806_05.jpg", "0000806_06.jpg", "0000806_07.jpg", "0000806_08.jpg", "0000806_09.jpg", "0000806_10.jpg", "0000806_11.jpg", "0000806_12.jpg", "0000806_13.jpg", "0000806_14.jpg", "0000806_15.jpg", "0000806_16.jpg", "0000806_17.jpg", "0000806_18.jpg", "0000806_19.jpg", "0000806_20.jpg", "0000806_21.jpg", "0000806_22.jpg", "0000806_23.jpg" ]

[ { "from": "human", "value": "\nSuppose you are an expert in judging and evaluating the quality of AI-generated videos,\nplease watch the following frames of a given video and see the text prompt for generating the video,\nthen give scores from 5 different dimensions:\n(1) visual quality: the quality of the video in terms of clearness, resolution, brightness, and color\n(2) temporal consistency, both the consistency of objects or humans and the smoothness of motion or movements\n(3) dynamic degree, the degree of dynamic changes\n(4) text-to-video alignment, the alignment between the text prompt and the video content\n(5) factual consistency, the consistency of the video content with the common-sense and factual knowledge\n\nfor each dimension, output a float number from 1.0 to 4.0,\nthe higher the number is, the better the video performs in that sub-score, \nthe lowest 1.0 means Bad, the highest 4.0 means Perfect/Real (the video is like a real video)\nHere is an output example:\nvisual quality: 3.2\ntemporal consistency: 2.7\ndynamic degree: 4.0\ntext-to-video alignment: 2.3\nfactual consistency: 1.8\n\nFor this video, the text prompt is \"70s atmospheric arthouse movie. A field of white moves menachingly, as if alive\",\nall the frames of video are as follows:\n\n<image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> " }, { "from": "gpt", "value": "visual quality: 3\n temporal consistency: 3\n dynamic degree: 2\n text-to-video alignment: 0\n factual consistency: 3\n" } ]

[ "8000035_00.jpg", "8000035_01.jpg", "8000035_02.jpg", "8000035_03.jpg", "8000035_04.jpg", "8000035_05.jpg", "8000035_06.jpg", "8000035_07.jpg", "8000035_08.jpg", "8000035_09.jpg", "8000035_10.jpg", "8000035_11.jpg", "8000035_12.jpg", "8000035_13.jpg", "8000035_14.jpg", "8000035_15.jpg", "8000035_16.jpg", "8000035_17.jpg", "8000035_18.jpg", "8000035_19.jpg", "8000035_20.jpg", "8000035_21.jpg", "8000035_22.jpg", "8000035_23.jpg" ]

[ { "from": "human", "value": "\nSuppose you are an expert in judging and evaluating the quality of AI-generated videos,\nplease watch the following frames of a given video and see the text prompt for generating the video,\nthen give scores from 5 different dimensions:\n(1) visual quality: the quality of the video in terms of clearness, resolution, brightness, and color\n(2) temporal consistency, both the consistency of objects or humans and the smoothness of motion or movements\n(3) dynamic degree, the degree of dynamic changes\n(4) text-to-video alignment, the alignment between the text prompt and the video content\n(5) factual consistency, the consistency of the video content with the common-sense and factual knowledge\n\nfor each dimension, output a float number from 1.0 to 4.0,\nthe higher the number is, the better the video performs in that sub-score, \nthe lowest 1.0 means Bad, the highest 4.0 means Perfect/Real (the video is like a real video)\nHere is an output example:\nvisual quality: 3.2\ntemporal consistency: 2.7\ndynamic degree: 4.0\ntext-to-video alignment: 2.3\nfactual consistency: 1.8\n\nFor this video, the text prompt is \"Start with a warm, inviting scene of a cozy family home 4k\",\nall the frames of video are as follows:\n\n<image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> " }, { "from": "gpt", "value": "visual quality: 2\n temporal consistency: 3\n dynamic degree: 3\n text-to-video alignment: 0\n factual consistency: 2\n" } ]

[ "8002982_00.jpg", "8002982_01.jpg", "8002982_02.jpg", "8002982_03.jpg", "8002982_04.jpg", "8002982_05.jpg", "8002982_06.jpg", "8002982_07.jpg", "8002982_08.jpg", "8002982_09.jpg", "8002982_10.jpg", "8002982_11.jpg", "8002982_12.jpg", "8002982_13.jpg", "8002982_14.jpg", "8002982_15.jpg", "8002982_16.jpg", "8002982_17.jpg", "8002982_18.jpg", "8002982_19.jpg", "8002982_20.jpg", "8002982_21.jpg", "8002982_22.jpg", "8002982_23.jpg" ]

[ { "from": "human", "value": "\nSuppose you are an expert in judging and evaluating the quality of AI-generated videos,\nplease watch the following frames of a given video and see the text prompt for generating the video,\nthen give scores from 5 different dimensions:\n(1) visual quality: the quality of the video in terms of clearness, resolution, brightness, and color\n(2) temporal consistency, both the consistency of objects or humans and the smoothness of motion or movements\n(3) dynamic degree, the degree of dynamic changes\n(4) text-to-video alignment, the alignment between the text prompt and the video content\n(5) factual consistency, the consistency of the video content with the common-sense and factual knowledge\n\nfor each dimension, output a float number from 1.0 to 4.0,\nthe higher the number is, the better the video performs in that sub-score, \nthe lowest 1.0 means Bad, the highest 4.0 means Perfect/Real (the video is like a real video)\nHere is an output example:\nvisual quality: 3.2\ntemporal consistency: 2.7\ndynamic degree: 4.0\ntext-to-video alignment: 2.3\nfactual consistency: 1.8\n\nFor this video, the text prompt is \"homer simpson walking ahead of an explosion\",\nall the frames of video are as follows:\n\n<image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> " }, { "from": "gpt", "value": "visual quality: 2\n temporal consistency: 1\n dynamic degree: 3\n text-to-video alignment: 0\n factual consistency: 1\n" } ]

[ "0001756_00.jpg", "0001756_01.jpg", "0001756_02.jpg", "0001756_03.jpg", "0001756_04.jpg", "0001756_05.jpg", "0001756_06.jpg", "0001756_07.jpg", "0001756_08.jpg", "0001756_09.jpg", "0001756_10.jpg", "0001756_11.jpg", "0001756_12.jpg", "0001756_13.jpg", "0001756_14.jpg", "0001756_15.jpg", "0001756_16.jpg", "0001756_17.jpg", "0001756_18.jpg", "0001756_19.jpg", "0001756_20.jpg", "0001756_21.jpg", "0001756_22.jpg", "0001756_23.jpg" ]

[ { "from": "human", "value": "\nSuppose you are an expert in judging and evaluating the quality of AI-generated videos,\nplease watch the following frames of a given video and see the text prompt for generating the video,\nthen give scores from 5 different dimensions:\n(1) visual quality: the quality of the video in terms of clearness, resolution, brightness, and color\n(2) temporal consistency, both the consistency of objects or humans and the smoothness of motion or movements\n(3) dynamic degree, the degree of dynamic changes\n(4) text-to-video alignment, the alignment between the text prompt and the video content\n(5) factual consistency, the consistency of the video content with the common-sense and factual knowledge\n\nfor each dimension, output a float number from 1.0 to 4.0,\nthe higher the number is, the better the video performs in that sub-score, \nthe lowest 1.0 means Bad, the highest 4.0 means Perfect/Real (the video is like a real video)\nHere is an output example:\nvisual quality: 3.2\ntemporal consistency: 2.7\ndynamic degree: 4.0\ntext-to-video alignment: 2.3\nfactual consistency: 1.8\n\nFor this video, the text prompt is \"the both girl and boy 20 year old in Japanese anime perform ritual again in garden\",\nall the frames of video are as follows:\n\n<image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> " }, { "from": "gpt", "value": "visual quality: 3\n temporal consistency: 3\n dynamic degree: 2\n text-to-video alignment: 0\n factual consistency: 3\n" } ]

[ "0002636_00.jpg", "0002636_01.jpg", "0002636_02.jpg", "0002636_03.jpg", "0002636_04.jpg", "0002636_05.jpg", "0002636_06.jpg", "0002636_07.jpg", "0002636_08.jpg", "0002636_09.jpg", "0002636_10.jpg", "0002636_11.jpg", "0002636_12.jpg", "0002636_13.jpg", "0002636_14.jpg", "0002636_15.jpg", "0002636_16.jpg", "0002636_17.jpg", "0002636_18.jpg", "0002636_19.jpg", "0002636_20.jpg", "0002636_21.jpg", "0002636_22.jpg", "0002636_23.jpg" ]

[ { "from": "human", "value": "\nSuppose you are an expert in judging and evaluating the quality of AI-generated videos,\nplease watch the following frames of a given video and see the text prompt for generating the video,\nthen give scores from 5 different dimensions:\n(1) visual quality: the quality of the video in terms of clearness, resolution, brightness, and color\n(2) temporal consistency, both the consistency of objects or humans and the smoothness of motion or movements\n(3) dynamic degree, the degree of dynamic changes\n(4) text-to-video alignment, the alignment between the text prompt and the video content\n(5) factual consistency, the consistency of the video content with the common-sense and factual knowledge\n\nfor each dimension, output a float number from 1.0 to 4.0,\nthe higher the number is, the better the video performs in that sub-score, \nthe lowest 1.0 means Bad, the highest 4.0 means Perfect/Real (the video is like a real video)\nHere is an output example:\nvisual quality: 3.2\ntemporal consistency: 2.7\ndynamic degree: 4.0\ntext-to-video alignment: 2.3\nfactual consistency: 1.8\n\nFor this video, the text prompt is \"There is a pure woman in the afternoon after the storm\",\nall the frames of video are as follows:\n\n<image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> " }, { "from": "gpt", "value": "visual quality: 3\n temporal consistency: 3\n dynamic degree: 1\n text-to-video alignment: 0\n factual consistency: 3\n" } ]