FlowMark: Mask-Guided Video Watermarking

2026-07-06

AI Summary

This work proposes the first end-to-end trainable adaptive video watermarking framework, addressing key limitations of existing methods that rely on manual masks and exhibit poor robustness under compression, temporal editing, and social media redistribution. The proposed approach integrates a mask prediction network to automatically identify optimal embedding regions, combined with region-aware encoding and noise-augmented training. This enables robust recovery against temporal manipulations—including frame swapping, insertion, and deletion—while eliminating flickering artifacts and preserving high visual quality (achieving a PSNR of 50.08 dB). Without requiring human-annotated masks, the method reliably embeds 128 bits of information and maintains effective traceability even after re-encoding on platforms such as YouTube and Facebook and under various distortions, significantly enhancing the practicality and robustness of video watermarking.

Abstract

We present FlowMark, a video watermarking framework guided by automatically predicted object masks. In contrast to prior region-based approaches that require user-supplied mask guidance, FlowMark learns to identify optimal regions for watermark embedding through a dedicated Mask Predictor network. Our end-to-end trainable architecture combines region-aware encoding with noise-augmented training to ensure robustness against compression, geometric transformations, and content variation, while preserving high perceptual quality. Our content-adaptive masking keeps watermark signals coherent with natural video dynamics, effectively eliminating perceptual flicker. Beyond compression robustness, FlowMark maintains reliable watermark recovery under video-native temporal edits (e.g., frame swap, insertion, deletion, resampling, and interpolation) and real-world social media distribution pipelines (e.g., YouTube and Facebook re-encoding). Experimental results on both image and video datasets show that FlowMark reliably embeds $128$-bit messages with up to $50.08$ dB PSNR, offering strong performance for content provenance, temporal authenticity verification, and video integrity protection.

Problem

• video watermarking
• robustness
• perceptual quality
• temporal authenticity
• content provenance

Innovation

• mask-guided watermarking
• content-adaptive masking
• temporal robustness
• end-to-end trainable framework
• video integrity protection