Explore research papers, interactive demos, and comprehensive documentation
| Repository | Description | Stars |
|---|---|---|
| Video Seal | Image & Video Watermarking |  |
| Audio Seal | Audio Watermarking |  |
| Text Seal | Text Watermarking |  |
| Dist Seal | In-Model Latent Space Watermarking |  |
| Stable Signature | Latent Diffusion Watermarking |  |
| WAM | Watermark Anything Model |  |
| WMAR | Autoregressive Image Generation |  |
Meta Seal is a comprehensive, open-source framework for invisible, robust watermarking across all modalities: audio, image, video, and text. This suite spans the entire generative AI lifecycle—from training data and inference to generated media—providing state-of-the-art tools for content provenance and authentication.
Watermarks applied after content generation by any model or system. Model-agnostic and universal across all content types.
| Model | Description | Resources |
|---|---|---|
| PixelSeal | 🏆 Flagship image & video watermarking model SOTA in terms of robustness and imperceptibility, built with a better and more stable adversarial-only training paradigm | Paper • Code |
| ChunkySeal | Bigger model with 4× capacity boost to 1024 bits while preserving quality and robustness | Paper • Code |
| VideoSeal | Extension of image watermarking models to video, resilient to editing and video codecs | Paper • Code • |
| WAM | Embed (possibly multiple) localized watermarks into images, survives inpainting and splicing attacks | Paper • Code |
| SyncSeal | Watermarking models for robust image synchronization, enabling to revert geometric transformations applied to image | Paper • Code |
| Model | Description | Resources |
|---|---|---|
| AudioSeal & AudioSeal Streaming | Localized audio watermarking with sample-level detection and streaming support for real-time applications | Paper • Code |
| Model | Description | Resources |
|---|---|---|
| TextSeal | Comprehensive evaluation framework for post-hoc text watermarking with LLM rephrasing | Paper • Code |
Watermarks embedded during content generation by modifying model behavior or latent representations.
| Model | Description | Resources |
|---|---|---|
| DISTSEAL | Unified latent space watermarking that enables 20× speedup over pixel methods and secures open-source models via in-model distillation | Paper • Code |
| Stable Signature | Roots the watermark in the model's latent decoder for tracing the outputs of latent generative models | Paper • Code |
| WMAR | Watermarking for autoregressive image generation models | Paper • Code |
Watermarks embedded into training datasets to track data provenance and detect unauthorized usage.
| Research | Description | Resources |
|---|---|---|
| Radioactive watermarks | Designed to detect if a language model was trained on synthetic text by detecting weak residuals of watermark signals in fine-tuned LLMs, with high confidence detection even when as little as 5% of training text is watermarked | Paper • Code |
| Detecting benchmark contamination through watermarking | Watermarks benchmarks before release to detect if models were trained on test sets, using theoretically grounded statistical tests to identify contamination while preserving benchmark utility | Paper • Code |
Research on adversarial attacks and defenses for watermarking systems through red teaming.
| Research | Description | Resources |
|---|---|---|
| WMForger | Black-box watermark forging using image preference models for red-teaming watermarking systems | Paper • Code |
The code is licensed under an MIT license.
