SPDMark: Selective Parameter Displacement for Robust Video Watermarking

2025-12-12

AI Summary

To address the challenge of simultaneously achieving imperceptibility, robustness, and efficiency in generative video provenance, this paper proposes a novel paradigm for embedding robust watermarks directly into the generation process of video diffusion models. Methodologically, we introduce an inter-layer basis vector displacement encoding scheme to enable indexable watermark embedding; integrate LoRA adapters with cryptographic hashing for frame-level precise localization; and impose temporal consistency constraints alongside joint optimization—balancing message recovery, perceptual similarity (LPIPS), and temporal coherence—to support recovery from frame-order tampering. Evaluated on both text-to-video and image-to-video diffusion models, our approach achieves watermark recovery accuracy exceeding 98% and LPIPS < 0.03, while demonstrating strong robustness against common attacks including compression, cropping, and frame insertion/deletion.

Technology Category

• Computer Vision: Diffusion Models for Vision
• Machine Learning: Adversarial Learning & Robustness
• Natural Language Processing: Generation

Application Category

• Security and Privacy: Data transparency and provenance
• User Modeling, Personalization and Recommendation: Attacks and countermeasures in recommendation systems
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking

Abstract

The advent of high-quality video generation models has amplified the need for robust watermarking schemes that can be used to reliably detect and track the provenance of generated videos. Existing video watermarking methods based on both post-hoc and in-generation approaches fail to simultaneously achieve imperceptibility, robustness, and computational efficiency. This work introduces a novel framework for in-generation video watermarking called SPDMark (pronounced `SpeedMark') based on selective parameter displacement of a video diffusion model. Watermarks are embedded into the generated videos by modifying a subset of parameters in the generative model. To make the problem tractable, the displacement is modeled as an additive composition of layer-wise basis shifts, where the final composition is indexed by the watermarking key. For parameter efficiency, this work specifically leverages low-rank adaptation (LoRA) to implement the basis shifts. During the training phase, the basis shifts and the watermark extractor are jointly learned by minimizing a combination of message recovery, perceptual similarity, and temporal consistency losses. To detect and localize temporal modifications in the watermarked videos, we use a cryptographic hashing function to derive frame-specific watermark messages from the given base watermarking key. During watermark extraction, maximum bipartite matching is applied to recover the correct frame order, even from temporally tampered videos. Evaluations on both text-to-video and image-to-video generation models demonstrate the ability of SPDMark to generate imperceptible watermarks that can be recovered with high accuracy and also establish its robustness against a variety of common video modifications.

Problem

• Robust watermarking for generated video provenance tracking
• Simultaneously achieving imperceptibility, robustness, and computational efficiency
• Detecting and localizing temporal modifications in watermarked videos

Innovation

• Selective parameter displacement in video diffusion models
• Low-rank adaptation for efficient watermark embedding
• Cryptographic hashing with bipartite matching for tamper detection