Dataset Preview
Full Screen
The full dataset viewer is not available (click to read why). Only showing a preview of the rows.
The dataset generation failed because of a cast error
Error code:   DatasetGenerationCastError
Exception:    DatasetGenerationCastError
Message:      An error occurred while generating the dataset

All the data files must have the same columns, but at some point there are 1 new columns ({'a man is on the ground while his car is parked next to him and he'}) and 3 missing columns ({'COCO_train2014_000000000009.jpg', 'plate rack broccoli', 'colorful dishes holding meat vegetables fruit and bread.'}).

This happened while the csv dataset builder was generating data using

zip://CC_caption_40.txt::/tmp/hf-datasets-cache/heavy/datasets/99234892092654-config-parquet-and-info-AhmedSSabir-Textual-Image-fb7117a2/downloads/818f97fe0a4486c2aeda74274d370fbbed88c8837783cf279774904438351c20

Please either edit the data files to have matching columns, or separate them into different configurations (see docs at https://hf.co/docs/hub/datasets-manual-configuration#multiple-configurations)
Traceback:    Traceback (most recent call last):
                File "/src/services/worker/.venv/lib/python3.9/site-packages/datasets/builder.py", line 2011, in _prepare_split_single
                  writer.write_table(table)
                File "/src/services/worker/.venv/lib/python3.9/site-packages/datasets/arrow_writer.py", line 585, in write_table
                  pa_table = table_cast(pa_table, self._schema)
                File "/src/services/worker/.venv/lib/python3.9/site-packages/datasets/table.py", line 2302, in table_cast
                  return cast_table_to_schema(table, schema)
                File "/src/services/worker/.venv/lib/python3.9/site-packages/datasets/table.py", line 2256, in cast_table_to_schema
                  raise CastError(
              datasets.table.CastError: Couldn't cast
              a man is on the ground while his car is parked next to him and he: string
              -- schema metadata --
              pandas: '{"index_columns": [{"kind": "range", "name": null, "start": 0, "' + 493
              to
              {'COCO_train2014_000000000009.jpg': Value(dtype='string', id=None), 'plate rack broccoli': Value(dtype='string', id=None), 'colorful dishes holding meat vegetables fruit and bread.': Value(dtype='string', id=None)}
              because column names don't match
              
              During handling of the above exception, another exception occurred:
              
              Traceback (most recent call last):
                File "/src/services/worker/src/worker/job_runners/config/parquet_and_info.py", line 1321, in compute_config_parquet_and_info_response
                  parquet_operations = convert_to_parquet(builder)
                File "/src/services/worker/src/worker/job_runners/config/parquet_and_info.py", line 935, in convert_to_parquet
                  builder.download_and_prepare(
                File "/src/services/worker/.venv/lib/python3.9/site-packages/datasets/builder.py", line 1027, in download_and_prepare
                  self._download_and_prepare(
                File "/src/services/worker/.venv/lib/python3.9/site-packages/datasets/builder.py", line 1122, in _download_and_prepare
                  self._prepare_split(split_generator, **prepare_split_kwargs)
                File "/src/services/worker/.venv/lib/python3.9/site-packages/datasets/builder.py", line 1882, in _prepare_split
                  for job_id, done, content in self._prepare_split_single(
                File "/src/services/worker/.venv/lib/python3.9/site-packages/datasets/builder.py", line 2013, in _prepare_split_single
                  raise DatasetGenerationCastError.from_cast_error(
              datasets.exceptions.DatasetGenerationCastError: An error occurred while generating the dataset
              
              All the data files must have the same columns, but at some point there are 1 new columns ({'a man is on the ground while his car is parked next to him and he'}) and 3 missing columns ({'COCO_train2014_000000000009.jpg', 'plate rack broccoli', 'colorful dishes holding meat vegetables fruit and bread.'}).
              
              This happened while the csv dataset builder was generating data using
              
              zip://CC_caption_40.txt::/tmp/hf-datasets-cache/heavy/datasets/99234892092654-config-parquet-and-info-AhmedSSabir-Textual-Image-fb7117a2/downloads/818f97fe0a4486c2aeda74274d370fbbed88c8837783cf279774904438351c20
              
              Please either edit the data files to have matching columns, or separate them into different configurations (see docs at https://hf.co/docs/hub/datasets-manual-configuration#multiple-configurations)

Need help to make the dataset viewer work? Make sure to review how to configure the dataset viewer, and open a discussion for direct support.

COCO_train2014_000000000009.jpg
string
plate rack broccoli
string
colorful dishes holding meat vegetables fruit and bread.
string
COCO_train2014_000000000025.jpg
gazelle cheetah
two giraffes standing in a tree filled area.
COCO_train2014_000000000030.jpg
vase flower
a white vase filled with different colored flowers.
COCO_train2014_000000000034.jpg
zebra plant
a lone zebra grazing in some green grass.
COCO_train2014_000000000036.jpg
umbrella dress human face
a woman with an umbrella near the sea
COCO_train2014_000000000049.jpg
pickelhaube cock horse
a pair of horses performing tricks in a field.
COCO_train2014_000000000061.jpg
siamang gorilla
there are people riding elephants in the middle of a forest
COCO_train2014_000000000064.jpg
parking meter analog clock car tree
a clock is seen in front of a tall tree.
COCO_train2014_000000000071.jpg
electric locomotive freight car
a train coming down the tracks in the city.
COCO_train2014_000000000072.jpg
upright whippet
two giraffes hang out near trees and nuzzle up to each other.
COCO_train2014_000000000077.jpg
corkscrew chain saw person footwear man
group of boys performing skateboard trick on ramp with graffiti on it
COCO_train2014_000000000078.jpg
table lamp analog clock
an antique looking clock beside an owl with a candle in it.
COCO_train2014_000000000081.jpg
airliner
a big plane with airfrance on the side of it.
COCO_train2014_000000000086.jpg
bicycle built for two tricycle
a person looks down at something while sitting on a bike.
COCO_train2014_000000000089.jpg
chiffonier microwave building countertop kitchen appliance
smooth top stove with exhaust fan that has light turned on.
COCO_train2014_000000000092.jpg
chocolate sauce
a chocolate desert on a plate with a fork.
COCO_train2014_000000000094.jpg
street sign
this is a view of a quaint city street.
COCO_train2014_000000000109.jpg
paddlewheel lakeside vehicle boat
a waterfront walkway and garden area next to a river.
COCO_train2014_000000000110.jpg
pizza
a woman cutting a pizza with a fork and knife.
COCO_train2014_000000000113.jpg
ballplayer jigsaw puzzle
two people standing near a table with a cake
COCO_train2014_000000000127.jpg
boathouse espresso
a slice of cake and mug sitting on a wooden table outside.
COCO_train2014_000000000138.jpg
chiffonier switch
a series of photographs of a tiny model kitchen
COCO_train2014_000000000142.jpg
skunk bagel dessert food drink
a close up of a plate of a jelly and banana sandwich
COCO_train2014_000000000144.jpg
upright fountain
two giraffes eating together out of a trough in a fenced area.
COCO_train2014_000000000149.jpg
kite park bench
a host of kites being flown on a field by people.
COCO_train2014_000000000151.jpg
freight car park bench
a man leans out of a vehicle near a short stop sign in a forest.
COCO_train2014_000000000154.jpg
zebra
three zebras standing on a grassy dirt field.
COCO_train2014_000000000165.jpg
military uniform vestment
a woman soldier holding up a pair of giant scissors to someones tie.
COCO_train2014_000000000194.jpg
pizza
a small personal sized pizza is shown on display.
COCO_train2014_000000000201.jpg
ski
a row of snow boards sticking out of the snow.
COCO_train2014_000000000247.jpg
plane wing
a red white and blue helicopter has a tag on one blade.
COCO_train2014_000000000250.jpg
street sign
street signs near tall buildings on the corner of greenwich st and vesey st.
COCO_train2014_000000000260.jpg
pay phone turnstile
the woman is standing with her luggage
COCO_train2014_000000000263.jpg
african elephant indian
a couple of elephants playing with each other in their pen.
COCO_train2014_000000000307.jpg
shetland sheepdog
a cute dog is in the grass with a frisbee.
COCO_train2014_000000000308.jpg
barrel caldron
a man pouring wine from casks for patrons
COCO_train2014_000000000309.jpg
cairn teddy
a small brown stuffed bear laying in the grass.
COCO_train2014_000000000312.jpg
african elephant
a herd of elephants including a tiny baby elephant cross a dirt road.
COCO_train2014_000000000315.jpg
sunscreen seashore
women are walking towards seats on the beach.
COCO_train2014_000000000321.jpg
hotdog
a sub sandwich in a box next to two hot dogs.
COCO_train2014_000000000322.jpg
projectile balance beam baseball bat person man
a person is throwing a frisbee at night.
COCO_train2014_000000000326.jpg
red wine bubble
a male in a white shirt is holding a glass of wine
COCO_train2014_000000000332.jpg
wok king crab food fast
asian noodles cut up egg roll with greens and carrots in a white bowl.
COCO_train2014_000000000349.jpg
electric locomotive
a white and orange train passing by trees and a lamppost.
COCO_train2014_000000000368.jpg
soccer ball
a young man kicking a soccer ball around a field.
COCO_train2014_000000000370.jpg
broccoli
a little girl holds a piece of broccoli towards the camera.
COCO_train2014_000000000382.jpg
ski alp person
a distance shot shows rolling snowy hills partially shadowed with ski prints reaching off into the distance and to the foreground a single skier.
COCO_train2014_000000000384.jpg
window shade
this is a living room with white curtains.
COCO_train2014_000000000389.jpg
bolo tie vestment man human face
the man smiles wearing a green neck tie near a crowd of people.
COCO_train2014_000000000394.jpg
sealyham terrier toy poodle
a white dog has a purple frisbee in its mouth.
COCO_train2014_000000000404.jpg
catamaran dock
three boats are docked together on the cloudy day.
COCO_train2014_000000000419.jpg
racket
a tennis player taking a swing at a ball
COCO_train2014_000000000431.jpg
racket
a male tennis player hits the ball on a grass court.
COCO_train2014_000000000436.jpg
harmonica bagel
a man licking his thumb with a peanut buttered bread slice in his hand.
COCO_train2014_000000000438.jpg
bakery pretzel dessert snack
a box full of donuts in different shapes and flavors.
COCO_train2014_000000000443.jpg
mitten cardigan
a person sitting on a couch with a cat laying on its back.
COCO_train2014_000000000446.jpg
bassoon joystick
a woman playing a video game indoors.
COCO_train2014_000000000450.jpg
pizza
a pizza sitting on a white plate next to a glass of water.
COCO_train2014_000000000471.jpg
school bus wheel
a yellow bus without passengers standing on the road.
COCO_train2014_000000000490.jpg
basset english foxhound person footwear building
dog and skateboard in a room and shadow of dog and person on a wall.
COCO_train2014_000000000491.jpg
eskimo dog wig
toy pomeranian dog on bag for playing with and enjoyment.
COCO_train2014_000000000508.jpg
beacon alp
a field covered in snow on top of a hillside.
COCO_train2014_000000000510.jpg
fountain park bench
an old man sits on a park bench beside a fountain
COCO_train2014_000000000514.jpg
theater curtain four poster
a very big and fancy looking room with a big pretty bed.
COCO_train2014_000000000529.jpg
bicycle built for two mountain bike motorcycle human face
a motorcyclist and a woman posing on a bike for a photo.
COCO_train2014_000000000531.jpg
tennis ball maze
two tennis courts both have doubles games playing on them.
COCO_train2014_000000000532.jpg
school bus fire engine
a man and two women disembark from two colorful busses.
COCO_train2014_000000000540.jpg
airliner
very large jal jet taxiing on the runway at airport in urban environment
COCO_train2014_000000000542.jpg
plane schooner helicopter wheel person
large group of people standing next to very old style air plane
COCO_train2014_000000000560.jpg
bathtub tub
this is a bathroom with a jacuzzi shower sink and toilet.
COCO_train2014_000000000562.jpg
pencil sharpener soup bowl
three colored toothbrushes standing in a glass holder.
COCO_train2014_000000000572.jpg
four poster crate
a person standing and a person looking down
COCO_train2014_000000000575.jpg
kit fox wallaby
a cat playing with a shoe in a grassy field.
COCO_train2014_000000000581.jpg
ibizan hound wallaby
there is a white and beige dog sitting on the floor
COCO_train2014_000000000584.jpg
acorn squash plate
a picture of a bowl of food that is mostly vegetables.
COCO_train2014_000000000595.jpg
rain barrel ashcan
an old tv monitor sits in the middle of a stream.
COCO_train2014_000000000597.jpg
african elephant
elephant herd in a line walking across savanna.
COCO_train2014_000000000605.jpg
partridge tray
a cup of liquid with a fancy design on top of it.
COCO_train2014_000000000612.jpg
theater curtain teddy
a group of teddy bears laying on top of a red bed.
COCO_train2014_000000000620.jpg
rotisserie safe
large cheese pizzas stacked in a warming oven.
COCO_train2014_000000000625.jpg
volleyball rugby ball
3 people attempt to catch a frisbee in midair.
COCO_train2014_000000000629.jpg
moped motor scooter
an old motorcycle rests on its kickstand by a door in front of a wall with a mural of a tree.
COCO_train2014_000000000634.jpg
corkscrew aircraft carrier
a man riding a skateboard on a ramp.
COCO_train2014_000000000643.jpg
desktop computer
a computer desk with a desktop computer and a stuffed turtle on top.
COCO_train2014_000000000650.jpg
madagascar cat tiger
cat sitting on top of car with brush behind.
COCO_train2014_000000000656.jpg
racer crash helmet
a motorcycle parked in a parking space in front of a store.
COCO_train2014_000000000659.jpg
electric locomotive school bus
a train engine carrying carts parked at a station.
COCO_train2014_000000000670.jpg
grocery store carton
a bunch of boxes full of assorted fruit on display.
COCO_train2014_000000000671.jpg
leonberg fountain camel tree
a statue of a soldier sitting atop a horse
COCO_train2014_000000000673.jpg
paddle boat vehicle
eleven surfboards face down on an empty beach.
COCO_train2014_000000000681.jpg
trolleybus
passenger bus parked on street with sunset in background
COCO_train2014_000000000684.jpg
car mirror hourglass
a three piece circular mirror with a girl taking a picture of it in the reflection
COCO_train2014_000000000690.jpg
corkscrew bicycle built for two tree land vehicle
a person riding a motorcycle on a road with trees
COCO_train2014_000000000706.jpg
upright cheetah giraffe tree
a giraffe standing in a stand of trees.
COCO_train2014_000000000714.jpg
washbasin
two different bathrooms one has a green toilet
COCO_train2014_000000000716.jpg
joystick laptop
the extra laptop is on standby for the computer users.
COCO_train2014_000000000722.jpg
kite parachute person land vehicle
clowns are on hand to assist fliers with their kites.
COCO_train2014_000000000723.jpg
restaurant
a large motorcycle sits outside of a bar.
COCO_train2014_000000000731.jpg
poncho umbrella
a woman holding her umbrella is standing in a quite large puddle.
COCO_train2014_000000000735.jpg
confectionery altar
a women cutting a large cake with one lit candle.
COCO_train2014_000000000753.jpg
hartebeest cheetah
the giraffe is standing alone in the wilderness.
End of preview.

Update: OCT-2023

Add v2 with recent SoTA model swinV2 classifier for both soft/hard-label visual_caption_cosine_score_v2 with person label (0.2, 0.3 and 0.4)

Introduction

Modern image captaining relies heavily on extracting knowledge, from images such as objects, to capture the concept of static story in the image. In this paper, we propose a textual visual context dataset for captioning, where the publicly available dataset COCO caption (Lin et al., 2014) has been extended with information about the scene (such as objects in the image). Since this information has textual form, it can be used to leverage any NLP task, such as text similarity or semantic relation methods, into captioning systems, either as an end-to-end training strategy or a post-processing based approach.

Please refer to project page and Github for more information. arXiv Website shields.io

For quick start please have a look this demo and pre-trained model with th 0.2, 0.3, 0.4

Overview

We enrich COCO-Caption with textual Visual Context information. We use ResNet152, CLIP, and Faster R-CNN to extract object information for each image. We use three filter approaches to ensure the quality of the dataset (1) Threshold: to filter out predictions where the object classifier is not confident enough, and (2) semantic alignment with semantic similarity to remove duplicated objects. (3) semantic relatedness score as soft-label: to guarantee the visual context and caption have a strong relation. In particular, we use Sentence-RoBERTa-sts via cosine similarity to give a soft score, and then we use a threshold to annotate the final label (if th ≥ 0.2, 0.3, 0.4 then 1,0). Finally, to take advantage of the visual overlap between caption and visual context, and to extract global information, we use BERT followed by a shallow 1D-CNN (Kim, 2014) to estimate the visual relatedness score.

Download

  1. Dowload Raw data with ID and Visual context -> original dataset with related ID caption train2014
  2. Downlod Data with cosine score-> soft cosine lable with th 0.2, 0.3, 0.4 and 0.5 and hardlabel [0,1]
  3. Dowload Overlaping visual with caption-> Overlap visual context and the human annotated caption
  4. Download Dataset (tsv file) 0.0-> raw data with hard lable without cosine similairty and with threshold cosine sim degree of the relation beteween the visual and caption = 0.2, 0.3, 0.4
  5. Download Dataset GenderBias-> man/woman replaced with person class label

Citation

The details of this repo are described in the following paper. If you find this repo useful, please kindly cite it:

@article{sabir2023visual,
  title={Visual Semantic Relatedness Dataset for Image Captioning},
  author={Sabir, Ahmed and Moreno-Noguer, Francesc and Padr{\'o}, Llu{\'\i}s},
  journal={arXiv preprint arXiv:2301.08784},
  year={2023}
}
Downloads last month
282