I2VWM: Robust Watermarking for Image to Video Generation

The rapid progress of image-guided video generation (I2V) has raised concerns about its potential misuse in misinformation and fraud, underscoring the urgent need for effective digital watermarking. While existing watermarking methods demonstrate robustness within a single modality, they fail to trace source images in I2V settings. To address this gap, we introduce the concept of Robust Diffusion Distance, which measures the temporal persistence of watermark signals in generated videos. Building on this, we propose I2VWM, a cross-modal watermarking framework designed to enhance watermark robustness across time. I2VWM leverages a video-simulation noise layer during training and employs an optical-flow-based alignment module during inference. Experiments on both open-source and commercial I2V models demonstrate that I2VWM significantly improves robustness while maintaining imperceptibility, establishing a new paradigm for cross-modal watermarking in the era of generative video. \href{https://github.com/MrCrims/I2VWM-Robust-Watermarking-for-Image-to-Video-Generation}{Code Released.}

Key Contributions

• Introduces Robust Diffusion Distance, a metric measuring temporal persistence of watermark signals across frames in I2V-generated videos
• Proposes I2VWM, a cross-modal watermarking framework using a video-simulation noise layer during training and optical-flow-based alignment at inference to maintain watermark robustness
• Demonstrates effectiveness on both open-source (Stable Video Diffusion, HunyuanVideo) and commercial I2V models

🛡️ Threat Analysis

Details

Similar Papers