Datasets:

Modalities:
Image
Text
Formats:
parquet
ArXiv:
Libraries:
Datasets
Dask
License:
mask
imagewidth (px)
128
1.6k
target_img_dataset
stringclasses
2 values
img_id
stringlengths
2
16
ann_id
stringlengths
1
8
COCO
339734
200890
COCO
537379
211233
COCO
109730
1991912
OpenImages
bcabc07fb161267a
c87485db
COCO
502857
636870
COCO
535467
1760742
COCO
182424
336326
COCO
32231
1635260
COCO
536540
61533
COCO
111492
148350
COCO
496855
603217
OpenImages
96a31c58ef46de06
41669b91
OpenImages
24463de8dd7e6225
5f64f48d
COCO
349400
1981591
COCO
472027
1271596
OpenImages
12cdc64468db41f6
95618870
OpenImages
a44292c4919d3500
5155dec4
COCO
53644
1081614
COCO
527060
1744350
COCO
404059
1184477
OpenImages
c50afd6612e78a70
3455b7b9
OpenImages
d461e61901e7587b
179e57c1
OpenImages
cc407b181731b811
292af2c8
OpenImages
d9b79379f94a8c5a
0556a89f
OpenImages
56ab651d77331e4c
ed8a7aa2
OpenImages
b18ce90fbf7a480c
ab0d8c95
OpenImages
3dcfbea09f80f319
0bb97127
OpenImages
5725caf6472131c9
ae4712d0
OpenImages
64735af0b6d1afdd
5ec8a2de
COCO
287094
571671
OpenImages
6a8cabf4c3c8d5f5
9a3656c3
OpenImages
a800cda6dc5133db
e38a4540
COCO
243341
44638
OpenImages
989681c5833c10c1
03946e16
OpenImages
f139ae1361d18bed
fc522af9
OpenImages
9359678320dba397
c4ea367d
OpenImages
406cbddea1e44ede
94b15a63
OpenImages
5f529edfac691407
f915a871
COCO
174386
1547411
OpenImages
6901d42354ef51e8
cfe5b610
OpenImages
50e24f711b7f64fc
4d32a564
OpenImages
942f5ae587a8250c
40869358
OpenImages
e426663f0fa2cc11
36228ed7
OpenImages
ebb02bca704c06c9
3313dddd
OpenImages
4bc1fb49d96743dc
0b2048b2
OpenImages
afc0d6105bd67821
865680ef
OpenImages
81fe210bbd86221f
e307fb28
OpenImages
cf872bd538f56d08
751d2867
COCO
74462
505390
OpenImages
f687c8326a19a0ae
c50c4fcf
OpenImages
78733aebec8862c5
634fcabd
OpenImages
190425ce5181e462
5150223b
COCO
20268
1668489
OpenImages
d96b4d0fe2dbcf9d
2f13e781
COCO
468443
248665
OpenImages
5e9577ec77fb1986
48c0ea76
OpenImages
d38e19fcf06bd478
18143866
OpenImages
6f1bb458eb81dd27
d7c42f5b
OpenImages
93c56b102008673b
fd22de9b
COCO
33631
690176
OpenImages
a824f3326c0882e0
05fc1882
COCO
303476
1151335
OpenImages
8ff0734feb171936
703e1918
OpenImages
10cd3b8a5a2e6453
3dc2af6b
OpenImages
32eba8346c7e3ea6
ee90a08f
OpenImages
7b362c303ac4952e
cb36d595
COCO
320246
1238482
OpenImages
c3d147bb51094f09
1c30c896
OpenImages
0f750eff98361262
b634978c
COCO
182902
1576916
OpenImages
300274bf2a9cdda9
d6ac2497
OpenImages
902e72bac053a974
d3ed1c8f
OpenImages
463179637f493c4b
358540e1
COCO
197036
1155245
OpenImages
8eae2ef70aed5b00
3d148173
OpenImages
db1a715ff5386f1a
6c0c097a
OpenImages
3dcd65fa72f5b1f8
bb016d2a
COCO
556695
1236056
COCO
418699
1159064
OpenImages
b1dae527887e680d
706e3a1c
OpenImages
ec02e6718e9cb470
4bbcae14
COCO
233954
31608
OpenImages
96077223ada88666
e709700c
OpenImages
b689aa91b2cdcc49
17a19de1
OpenImages
bad221bc6a40f4e3
92173762
COCO
48685
455840
OpenImages
66c62d782573ea4d
e641bfcf
COCO
305564
1116941
OpenImages
f03c1eb257788c94
c3010665
OpenImages
7bffe5e4bb8445ff
f1280283
OpenImages
9902931244c9d83e
c1fc319e
COCO
264322
1352560
COCO
225241
1298142
COCO
297299
86104
OpenImages
eb745891cce328f9
6b9f8bcc
OpenImages
d2fe2358980da9b2
d80ce8f8
OpenImages
9f4b0c2435246174
2a3e169b
OpenImages
bba9fb0465b5b436
7a16611b
OpenImages
ff2686ab2649e866
262f7e2a
OpenImages
b90f485f9d27cd9d
b0976715

Dataset Card for PIPE Masks Dataset

Dataset Summary

The PIPE (Paint by InPaint Edit) dataset is designed to enhance the efficacy of mask-free, instruction-following image editing models by providing a large-scale collection of image pairs and diverse object addition instructions. Here, we provide the masks used for the inpainting process to generate the source image for the PIPE dataset for both the train and test sets. Further details can be found in our project page and paper.

Columns

  • mask: The removed object mask used for creating the inpainted image.
  • target_img_dataset: The dataset to which the target image belongs.
  • img_id: The unique identifier of the GT image (the target image).
  • ann_id: The identifier of the object segmentation annotation of the object removed.

Loading the PIPE Masks Dataset

Here is an example of how to load and use this dataset with the datasets library:

from datasets import load_dataset

data_files = {"train": "data/train-*", "test": "data/test-*"}
dataset_masks  = load_dataset('paint-by-inpaint/PIPE_Masks',data_files=data_files)

# Display an example
example_train_mask = dataset_masks['train'][0]
print(example_train_mask)

example_test_mask = dataset_masks['test'][0]
print(example_test_mask)
Downloads last month
123