🎬 TimeSformer Violence Detection System

Upload a video to detect the presence of violent content using a state-of-the-art TimeSformer model.

How it works:

1. Upload a video file
2. The model samples 16 frames uniformly from the video
3. TimeSformer analyzes spatial and temporal patterns
4. Returns prediction with confidence scores

Output:

• predicted_class: Violence or Non-Violence
• confidence: Confidence score (0-1)
• frames_analyzed: Number of frames processed