Search is not available for this dataset
image
imagewidth (px)
512
512
bbox
sequencelengths
4
4
id
int64
0
12.8k
area
int64
198
93.6k
image_id
int64
158
582k
category_id
int64
1
2
[ 255, 255, 203, 72 ]
9,738
14,616
158
1
[ 412, 438, 148, 26 ]
4,285
3,848
166
2
[ 256, 256, 237, 237 ]
7,704
56,169
179
1
[ 336, 388, 308, 61 ]
7,071
18,788
186
2
[ 129, 424, 209, 111 ]
3,414
23,199
211
1
[ 266, 373, 445, 71 ]
10,711
31,595
292
2
[ 323, 149, 335, 67 ]
2,671
22,445
346
2
[ 335, 357, 302, 257 ]
9,928
77,614
398
1
[ 107, 40, 165, 30 ]
6,280
4,950
402
1
[ 377, 450, 218, 71 ]
12,510
15,478
425
1
[ 250, 254, 263, 85 ]
2,046
22,355
433
1
[ 94, 421, 136, 43 ]
6,433
5,848
465
2
[ 165, 75, 281, 100 ]
6,004
28,100
506
1
[ 105, 448, 160, 76 ]
6,616
12,160
579
1
[ 188, 130, 322, 51 ]
6,377
16,422
602
2
[ 347, 407, 268, 41 ]
5,031
10,988
691
2
[ 250, 238, 269, 245 ]
11,242
65,905
740
1
[ 124, 439, 199, 93 ]
8,804
18,507
770
1
[ 258, 263, 415, 72 ]
9,623
29,880
791
2
[ 253, 254, 276, 51 ]
1,260
14,076
854
1
[ 319, 122, 331, 59 ]
7,365
19,529
895
2
[ 255, 258, 269, 54 ]
10,457
14,526
903
2
[ 342, 449, 283, 73 ]
11,842
20,659
1,055
1
[ 115, 47, 180, 44 ]
3,254
7,920
1,093
1
[ 157, 413, 252, 39 ]
10,541
9,828
1,123
2
[ 345, 67, 274, 85 ]
7,140
23,290
1,182
1
[ 167, 415, 276, 37 ]
6,725
10,212
1,202
2
[ 142, 54, 234, 59 ]
12,028
13,806
1,234
1
[ 203, 374, 351, 62 ]
4,183
21,762
1,247
2
[ 256, 256, 297, 84 ]
4,592
24,948
1,288
1
[ 144, 105, 234, 51 ]
4,559
11,934
1,302
2
[ 176, 85, 303, 120 ]
4,018
36,360
1,303
1
[ 256, 255, 254, 47 ]
5,986
11,938
1,334
1
[ 246, 350, 441, 74 ]
3,273
32,634
1,356
2
[ 256, 243, 266, 67 ]
3,285
17,822
1,484
2
[ 168, 172, 286, 294 ]
3,600
84,084
1,491
1
[ 256, 256, 255, 145 ]
7,518
36,975
1,495
1
[ 271, 176, 481, 110 ]
8,223
52,910
1,571
2
[ 391, 53, 176, 57 ]
8,030
10,032
1,686
1
[ 339, 383, 283, 65 ]
11,101
18,395
1,726
2
[ 307, 150, 351, 77 ]
11,365
27,027
1,738
2
[ 352, 97, 262, 40 ]
343
10,480
1,753
2
[ 257, 158, 446, 78 ]
7,171
34,788
1,788
2
[ 150, 85, 250, 121 ]
11,976
30,250
1,850
1
[ 255, 261, 278, 49 ]
1,119
13,622
1,878
2
[ 257, 257, 122, 21 ]
4,417
2,562
1,970
2
[ 345, 104, 274, 43 ]
8,307
11,782
1,983
2
[ 256, 355, 442, 69 ]
3,893
30,498
2,031
2
[ 257, 245, 246, 62 ]
1,969
15,252
2,097
2
[ 350, 139, 250, 228 ]
682
57,000
2,143
1
[ 385, 472, 195, 21 ]
7,810
4,095
2,158
1
[ 151, 352, 253, 256 ]
10,498
64,768
2,336
1
[ 367, 422, 231, 34 ]
3,371
7,854
2,378
2
[ 303, 129, 351, 54 ]
4,259
18,954
2,387
2
[ 107, 437, 160, 30 ]
11,587
4,800
2,394
2
[ 338, 90, 292, 130 ]
3,404
37,960
2,398
1
[ 381, 427, 183, 31 ]
10,331
5,673
2,486
2
[ 382, 82, 216, 38 ]
4,265
8,208
2,494
2
[ 267, 352, 481, 74 ]
7,029
35,594
2,548
2
[ 155, 125, 255, 62 ]
649
15,810
2,561
2
[ 119, 441, 189, 84 ]
10,740
15,876
2,573
1
[ 307, 368, 348, 79 ]
3,981
27,492
2,631
2
[ 257, 252, 204, 31 ]
2,903
6,324
2,671
2
[ 360, 416, 245, 37 ]
5,839
9,065
2,716
2
[ 221, 153, 390, 86 ]
1,169
33,540
2,726
2
[ 359, 151, 253, 253 ]
7,930
64,009
2,775
1
[ 208, 387, 354, 47 ]
5,629
16,638
2,807
2
[ 379, 419, 214, 134 ]
5,678
28,676
2,844
1
[ 242, 256, 218, 219 ]
12,762
47,742
2,855
1
[ 399, 334, 147, 291 ]
5,770
42,777
3,078
1
[ 255, 262, 377, 70 ]
1,125
26,390
3,088
2
[ 255, 258, 393, 61 ]
8,227
23,973
3,133
2
[ 232, 143, 397, 62 ]
8,029
24,614
3,158
2
[ 256, 255, 206, 52 ]
4,904
10,712
3,165
1
[ 173, 71, 297, 92 ]
804
27,324
3,175
1
[ 353, 155, 261, 261 ]
7,608
68,121
3,235
1
[ 341, 107, 290, 165 ]
8,574
47,850
3,274
1
[ 150, 462, 250, 39 ]
826
9,750
3,277
1
[ 322, 361, 330, 80 ]
2,781
26,400
3,281
2
[ 167, 96, 281, 39 ]
5,549
10,959
3,303
2
[ 420, 68, 132, 22 ]
9,673
2,904
3,350
2
[ 124, 420, 199, 122 ]
11,424
24,278
3,416
1
[ 172, 86, 295, 122 ]
7,300
35,990
3,463
1
[ 161, 434, 273, 93 ]
10,010
25,389
3,489
1
[ 350, 160, 271, 271 ]
5,958
73,441
3,490
1
[ 256, 254, 305, 48 ]
4,081
14,640
3,503
2
[ 154, 340, 258, 250 ]
5,810
64,500
3,507
1
[ 255, 252, 239, 63 ]
9,485
15,057
3,523
2
[ 370, 50, 228, 50 ]
8,842
11,400
3,524
1
[ 255, 256, 198, 51 ]
5,506
10,098
3,525
1
[ 256, 255, 216, 81 ]
1,572
17,496
3,543
1
[ 350, 57, 268, 64 ]
3,308
17,152
3,691
1
[ 250, 364, 469, 70 ]
9,005
32,830
3,697
2
[ 252, 351, 455, 88 ]
3,567
40,040
3,758
2
[ 394, 450, 174, 72 ]
5,694
12,528
3,809
1
[ 254, 255, 205, 22 ]
1,474
4,510
3,818
1
[ 122, 54, 194, 58 ]
2,908
11,252
3,821
1
[ 256, 256, 223, 86 ]
4,156
19,178
3,879
1
[ 125, 361, 44, 63 ]
10,753
2,772
3,908
2
[ 114, 443, 178, 86 ]
1,416
15,308
3,940
1

Visible watermarks datasets

We have observed that while datasets such as COCO are available for object detection, the availability of datasets specifically designed for the detection of watermarks added to images is significantly limited. Through our research, we identified only one such dataset, which originates from the paper Wdnet: Watermark-Decomposition Network for Visible Watermark Removal [1]. This dataset provides a collection of images along with their corresponding watermark masks for the purpose of watermark removal. Additionally, we noted that accessing this dataset presented challenges in terms of data accessibility and regeneration of dataset samples.

The CLWD Dataset, introduced in Wdnet: Watermark-Decomposition Network for Visible Watermark Removal [1], comprises images sourced from the COCO Dataset (Lin et al., 2014) [ 2] and masks of colored watermarks featuring random positions and opacities

Dataset Details (PITA Dataset)

We decided to introduce the Pita dataset, which is based on images from the COCO dataset (Lin et al., 2014) [ 2] and combines these with logos from the Open Logo Detection Challenge (Su et al., 2018) [3]. The dataset introduces several changes compared to other datasets, with a focus on the task of watermark detection rather than watermark removal.

The dataset is structured into three splits: a training split, a validation split, and a test split, collectively comprising approximately 20 000 watermarked images featuring both logos and text. We decided to incorporate two types of labels:

  • Text: The images are watermarked with a random font available on the computer used for generation, and the text size is also randomized.
  • Logos: The logos are sourced from the Open Logo Detection Challenge dataset (Su et al., 2018) and are characterized by random sizes and opacities.

The position of the logo or text is randomly selected from a set of available positions, specifically corners or the center. This restriction was introduced based on the observation that watermarks on social media or stock image websites are predominantly located in these positions.

The dataset is accompanied by command-line interface tools that facilitate reproducibility. These tools support both YOLO and Hugging Face formats, allowing the download of the dataset and generation with ease

Dataset Sources

Uses

  • DETR with Hugging face Transformers
  • YoloV8 model with ultralytics
  • FastRCNN with Pytorch Lighthning

Source Data

  • COCO dataset (Lin et al., 2014) [ 2] and
  • Open Logo Detection Challenge (Su et al., 2018) [3].

Data Collection and Processing

Generation of the dataset is reproducible using the cli tool of this repository.

A --help option is available in order to describe how to use the tool.

Annotation process

Logo were added to COCO images by applying rotation, scaling and opacity changes at a random position on the image. on the image.

Downloads last month
114
Edit dataset card

Models trained or fine-tuned on bastienp/visible-watermark-pita