linkedlist771/SoraWatermarkCleaner

Health v2 · maintenance only

Flags: archived

Adoption not part of the score

1149 stars · 241 forks observed · 2026-08-28

What it is AI-extracted, prompt v1, taxonomy v1, 2026-08-30, confidence not recorded

A deep learning tool that detects and removes the Sora2 watermark from AI-generated videos using a YOLO-based detector plus a restoration model, with batch processing and time-consistent results. The project has been archived since OpenAI discontinued the Sora model, with a successor project (DeMark-World) for other video generators.

Use cases

• remove sora watermark from ai generated video
• clean sora2 videos locally with gpu
• batch remove watermarks from generated videos
• detect watermark position in video frames
• train a custom watermark detection model
• remove watermark without flicker preserving time consistency

When to choose

• you need to remove the Sora2 watermark from videos offline with high quality
• you want a self-hosted GPU pipeline with batch processing
• you want to train or improve a watermark detector using the released labeled dataset

When to avoid

• you need to remove watermarks from other generators like Veo or Runway - use the successor DeMark-World instead
• you need a maintained tool - the project is archived and unmaintained
• you lack a CUDA GPU - the Docker image is ~20 GB and requires NVIDIA hardware
• removing watermarks from videos you do not have rights to

Facets

application · maturity abandoned

video-processing computer-vision machine-learning deep-learning computer-vision machine-learning media python windows cross-platform watermark-removal sora2 yolo object-detection video-inpainting archived-project ai-generated-video video docker gpu

Member repositories

For agents

markdown · JSON · MCP: product_card(name="linkedlist771/SoraWatermarkCleaner")