DamonDemon commited on
Commit
fed9745
1 Parent(s): e873158
Files changed (1) hide show
  1. src/about.py +4 -4
src/about.py CHANGED
@@ -29,21 +29,21 @@ SUB_TITLE = """<h2 align="center" id="space-title">Effective and efficient adver
29
  # What does your leaderboard evaluate?
30
  INTRODUCTION_TEXT = """
31
  This benchmark evaluates the <strong>robustness and utility retaining</strong> of safety-driven unlearned diffusion models (DMs) across a variety of tasks. For more details, please visit the [project](https://www.optml-group.com/posts/mu_attack).
32
- - The robustness of unlearned DM is evaluated through our proposed adversarial prompt attack, [UnlearnDiffAtk](https://github.com/OPTML-Group/Diffusion-MU-Attack), which has been accepted to ECCV 2024.
33
- - The utility retaining of unlearned DM is evaluated through FID and CLIP score on the generated images using [10K randomly sampled COCO caption prompts](https://github.com/OPTML-Group/Diffusion-MU-Attack/blob/main/prompts/coco_10k.csv).
34
 
35
  Demo of our offensive method: [UnlearnDiffAtk](https://huggingface.co/spaces/Intel/UnlearnDiffAtk)\\
36
  Demo of our defensive method: [AdvUnlearn](https://huggingface.co/spaces/Intel/AdvUnlearn)
37
  """
38
 
39
  EVALUATION_QUEUE_TEXT = """
40
- <strong>Evaluation Metrics</strong>:
41
  - Pre-Attack Success Rate (<strong>Pre-ASR</strong>): lower is better;
42
  - Post-attack success rate (<strong>Post-ASR</strong>): lower is better;
43
  - Fréchet inception distance(<strong>FID</strong>): evaluate distributional quality of image generations, lower is better;
44
  - <strong>CLIP Score</strong>: measure contextual alignment with prompt descriptions, higher is better.
45
 
46
- <strong>DM Unlearning Tasks</strong>:
47
  - NSFW: Nudity
48
  - Style: Van Gogh
49
  - Objects: Church, Tench, Parachute, Garbage Truck
 
29
  # What does your leaderboard evaluate?
30
  INTRODUCTION_TEXT = """
31
  This benchmark evaluates the <strong>robustness and utility retaining</strong> of safety-driven unlearned diffusion models (DMs) across a variety of tasks. For more details, please visit the [project](https://www.optml-group.com/posts/mu_attack).
32
+ - The <strong>robustness</strong> of unlearned DM is evaluated through our proposed adversarial prompt attack, [UnlearnDiffAtk](https://github.com/OPTML-Group/Diffusion-MU-Attack), which has been accepted to ECCV 2024.
33
+ - The <strong>utility retaining</strong> of unlearned DM is evaluated through FID and CLIP score on the generated images using [10K randomly sampled COCO caption prompts](https://github.com/OPTML-Group/Diffusion-MU-Attack/blob/main/prompts/coco_10k.csv).
34
 
35
  Demo of our offensive method: [UnlearnDiffAtk](https://huggingface.co/spaces/Intel/UnlearnDiffAtk)\\
36
  Demo of our defensive method: [AdvUnlearn](https://huggingface.co/spaces/Intel/AdvUnlearn)
37
  """
38
 
39
  EVALUATION_QUEUE_TEXT = """
40
+ <strong>\[Evaluation Metrics\]</strong>:
41
  - Pre-Attack Success Rate (<strong>Pre-ASR</strong>): lower is better;
42
  - Post-attack success rate (<strong>Post-ASR</strong>): lower is better;
43
  - Fréchet inception distance(<strong>FID</strong>): evaluate distributional quality of image generations, lower is better;
44
  - <strong>CLIP Score</strong>: measure contextual alignment with prompt descriptions, higher is better.
45
 
46
+ <strong>\[DM Unlearning Tasks\]</strong>:
47
  - NSFW: Nudity
48
  - Style: Van Gogh
49
  - Objects: Church, Tench, Parachute, Garbage Truck