Upload app.py with huggingface_hub

app.py

@@ -108,21 +108,20 @@ with gr.Blocks(css="""
     gr.Markdown("""
 # DFBench: The Image Deepfake Detection Benchmark 2025
 
-DFBench provides a standardized evaluation framework for computer vision deepfake detection systems.
+DFBench provides a standardized evaluation for computer vision deepfake detection systems.
 This leaderboard focuses on image deepfake detection, e.g. the output of text-to-image and image-to-image models.
 
 **Objectives:**
-- Allow fair comparison between deepfake detection models on unseen test data 
+- Allow fair comparison between deepfake detection models on unseen test data (no fine tuning on the test data possible)
 - Advance the state-of-the-art in synthetic media identification
 
-This benchmark serves the academic and industry research community by providing consistent evaluation standards for deepfake detection methodologies.
 """)
 
     with gr.Tab("Leaderboard"):
         gr.Markdown("## Leaderboard Image Deepfake Detection")
         gr.HTML(leaderboard_view())
         gr.Markdown("""
-*The Leaderboard is updated upon validation of new submissions. All results are evaluated using standardized metrics on the official [test dataset](https://huggingface.co/datasets/DFBench/Image-Deepfake-Detection-25).*
+*The Leaderboard is updated upon validation of new submissions. All results are evaluated on the official [test dataset](https://huggingface.co/datasets/DFBench/Image-Deepfake-Detection-25).*
         """)
 
     with gr.Tab("Submission Guidelines"):
@@ -134,7 +133,7 @@ This benchmark serves the academic and industry research community by providing
 The test dataset comprises **2,920 images**. The test data is unlabeled. Each image is either:
 - **Real:** An authentic, unmodified image
 - **Fake:** AI-generated or synthetically modified content
-
+Since there are no labels, you cannot (and should not) train your model on the test data.
 ---
 
 ## Submission Requirements
@@ -157,15 +156,13 @@ filename,label
 ### Submission Process
 
 1. Generate predictions for all 2,920 test images
-2. Format results according to specification above
+2. Format results according to specification above 
 3. Send your CSV file submission to: **submission@df-bench.com**. The name of the file should correspond to the leaderboard model name, e.g. `Model_This_name.csv` will be included as `Model This name` in the leaderboard.
 
 ### Evaluation Timeline
 - Submissions are processed within 5-7 business days
 - Approved submissions are added to the public leaderboard
 
----
-
 ## Notes
 - Each research group may submit one set of scores per month
 - All submissions undergo automated validation before leaderboard inclusion