Update README.md
Browse files
README.md
CHANGED
@@ -96,7 +96,7 @@ A data point comprises an image and its face and license plate annotations.
|
|
96 |
`width = <absolute_width> / <image_width>`
|
97 |
- `height`: normalized wheightdth of the bounding box.
|
98 |
`height = <absolute_height> / <image_height>`
|
99 |
-
- Example lines in YOLO v1.1 format `.txt' annotation file:
|
100 |
` 1 0.716797 0.395833 0.216406 0.147222
|
101 |
0 0.687109 0.379167 0.255469 0.158333
|
102 |
1 0.420312 0.395833 0.140625 0.166667
|
@@ -109,13 +109,17 @@ A data point comprises an image and its face and license plate annotations.
|
|
109 |
#### Initial Data Collection and Normalization
|
110 |
|
111 |
The objective of PP4AV is to build a benchmark dataset that can be used to evaluate face and license plate detection models for autonomous driving. For normal camera data, we sampled images from the existing videos in which cameras were mounted in moving vehicles, running around the European cities. We focus on sampling data in urban areas rather than highways in order to provide sufficient samples of license plates and pedestrians. The images in PP4AV were sampled from **6** European cities at various times of day, including nighttime. The source data from 6 cities in European was described as follow:
|
112 |
-
- `Paris`:
|
113 |
-
|
114 |
-
- `Netherland
|
115 |
-
|
116 |
-
- `
|
117 |
-
|
118 |
-
- `
|
|
|
|
|
|
|
|
|
119 |
|
120 |
We use the fisheye images from the WoodScape dataset to select **244** images from the front, rear, left, and right cameras for fisheye camera data.
|
121 |
The source of fisheye data for sampling is located at WoodScape's [fisheye images](https://woodscape.valeo.com/download).
|
|
|
96 |
`width = <absolute_width> / <image_width>`
|
97 |
- `height`: normalized wheightdth of the bounding box.
|
98 |
`height = <absolute_height> / <image_height>`
|
99 |
+
- Example lines in YOLO v1.1 format `.txt' annotation file:
|
100 |
` 1 0.716797 0.395833 0.216406 0.147222
|
101 |
0 0.687109 0.379167 0.255469 0.158333
|
102 |
1 0.420312 0.395833 0.140625 0.166667
|
|
|
109 |
#### Initial Data Collection and Normalization
|
110 |
|
111 |
The objective of PP4AV is to build a benchmark dataset that can be used to evaluate face and license plate detection models for autonomous driving. For normal camera data, we sampled images from the existing videos in which cameras were mounted in moving vehicles, running around the European cities. We focus on sampling data in urban areas rather than highways in order to provide sufficient samples of license plates and pedestrians. The images in PP4AV were sampled from **6** European cities at various times of day, including nighttime. The source data from 6 cities in European was described as follow:
|
112 |
+
- `Paris`: This subset contains **1450** images of the car driving down a Parisian street during the day:
|
113 |
+
URL: [paris_youtube_video](https://www.youtube.com/watch?v=nqWtGWymV6c)
|
114 |
+
- `Netherland day time`: This subset consists of **388** images of Hague, Amsterdam city in day time sampled by the following video:
|
115 |
+
[netherland_youtube_video]()
|
116 |
+
- `Netherland night time`: This subset consists of **824** images of Hague, Amsterdam city in night time sampled by the following video:
|
117 |
+
[netherland_youtube_video]()
|
118 |
+
- `Switzerland`: This subset consists of **372** images of Switzerland sampled by the following video:
|
119 |
+
[switzerland_youtube_video](https://www.youtube.com/watch?v=0iw5IP94m0Q)
|
120 |
+
- `Zurich`: This subset consists of 50 images of Zurich city provided by the Cityscapes training set in [leftImg8bit_trainvaltest.zip package](https://www.cityscapes-dataset.com/file-handling/?packageID=3)
|
121 |
+
- `Stuttgart`: This subset consists of 69 images of Stuttgart city provided by the Cityscapes training set in [leftImg8bit_trainvaltest.zip package](https://www.cityscapes-dataset.com/file-handling/?packageID=3)
|
122 |
+
- `Strasbourg`: This subset consists of 50 images of Strasbourg city provided by the Cityscapes training set in [leftImg8bit_trainvaltest.zip package](https://www.cityscapes-dataset.com/file-handling/?packageID=3)
|
123 |
|
124 |
We use the fisheye images from the WoodScape dataset to select **244** images from the front, rear, left, and right cameras for fisheye camera data.
|
125 |
The source of fisheye data for sampling is located at WoodScape's [fisheye images](https://woodscape.valeo.com/download).
|