Video Seal
Invisible video watermarks that survive blur, crops and compression
Latest: Video Seal (Dec 2024)
An open framework and state-of-the-art model for invisible neural video watermarking (December 2024), aimed at AI-content provenance: temporal watermark propagation converts any image watermarker into an efficient video one without stamping every high-res frame. MIT-licensed with training code, baselines, and a public demo.
Why it matters
Video Seal made invisible neural video watermarking open and efficient: temporal propagation turns any image watermarker into a video one without stamping every high-res frame, and marks survive blur, crops and compression. Released MIT-licensed as deepfake concerns peaked, it is the provenance counterpart to Meta's generative video push.
Facts
- Watermarks survive blurring, cropping, and compression edits; released MIT-licensed as deepfake concerns peaked — Meta's provenance counterpart to its generative video push.
Try it yourself
Official Video Seal demo ↗ pip install videoseal ↗ Run the official notebook in Colab ↗
Sources
arXiv ↗GitHub · videoseal ↗Live demo ↗