Datasets:
Update README.md
Browse files
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
|
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 |
|