Datasets:

Modalities:
Image
Languages:
English
ArXiv:
Libraries:
Datasets
License:
Shilin-LU commited on
Commit
c81f392
·
verified ·
1 Parent(s): a13199b

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +2 -2
README.md CHANGED
@@ -12,7 +12,7 @@ size: 10,000 instances
12
 
13
  **W-Bench is the first benchmark to evaluate watermarking robustness across four types of image editing techniques, including regeneration, global editing, local editing, and image-to-video generation.**
14
 
15
- Eleven representative watermarking methods are evaluated on the W-Bench. The W-Bench contains 10,000 samples sourced from datasets such as COCO, Flickr, ShareGPT4V, etc.
16
 
17
  GitHub Repo: [https://github.com/Shilin-LU/VINE](https://github.com/Shilin-LU/VINE)
18
 
@@ -24,7 +24,7 @@ The evaluation set consists of six subsets, each targeting a different type of A
24
  - 1,000 samples for global editing
25
  - 5,000 samples for local editing (divided into five sets, each containing 1,000 images and 1,000 masks, with different mask sizes ranging from 10–60% of the image area)
26
  - 1,000 samples for image-to-video generation
27
- - 1,000 samples for testing conventional distortion
28
 
29
  # How to download and use 🍷 W-Bench
30
 
 
12
 
13
  **W-Bench is the first benchmark to evaluate watermarking robustness across four types of image editing techniques, including regeneration, global editing, local editing, and image-to-video generation.**
14
 
15
+ Eleven representative watermarking methods are evaluated on the W-Bench. The W-Bench contains 10,000 instances sourced from datasets such as COCO, Flickr, ShareGPT4V, etc.
16
 
17
  GitHub Repo: [https://github.com/Shilin-LU/VINE](https://github.com/Shilin-LU/VINE)
18
 
 
24
  - 1,000 samples for global editing
25
  - 5,000 samples for local editing (divided into five sets, each containing 1,000 images and 1,000 masks, with different mask sizes ranging from 10–60% of the image area)
26
  - 1,000 samples for image-to-video generation
27
+ - 1,000 samples for testing conventional distortion (identical to the 1,000 samples used for deterministic regeneration)
28
 
29
  # How to download and use 🍷 W-Bench
30