GIFGuard: Proactive Forensics against Deepfakes in Facial GIFs via Spatiotemporal Watermarking

TL;DR

GIFGuard introduces a novel spatiotemporal watermarking framework for proactive deepfake detection in GIFs, addressing the limitations of static-image methods and ensuring robustness against facial manipulations.

Contribution

The paper presents the first tailored spatiotemporal watermarking framework for GIFs, including new encoding and decoding architectures, and introduces GIFfaces, a large-scale benchmark dataset for this domain.

Findings

GIFGuard achieves high visual fidelity and robustness against deepfakes.

The proposed methods outperform existing static-image forensic techniques.

GIFfaces facilitates future research in GIF deepfake forensics.

Abstract

The rapid evolution of deepfake technology poses an unprecedented threat to the authenticity of Graphics Interchange Format (GIF) imagery, which serves as a representative of short-loop temporal media in social networks. However, existing proactive forensics works are designed for static images, which limits their applicability to animated GIFs. To bridge this gap, we propose GIFGuard, the first spatiotemporal watermarking framework tailored for deepfake proactive forensics in GIFs. In the embedding stage, we propose the Spatiotemporal Adaptive Residual Encoder (STARE) to ensure robustness against high-level semantic tampering. It employs a 3D convolutional backbone with adaptive channel recalibration to capture globally coherent temporal dependencies. In the extraction stage, we design the Deep Integrity Restoration Decoder (DIRD). It utilizes a spatiotemporal hourglass architecture equipped with 3D attention to restore latent features, allowing for the accurate extraction of watermark signals even under severe facial manipulation. Furthermore, we construct GIFfaces, the first large-scale benchmark dataset curated for GIF proactive forensics to facilitate research in this domain. Extensive results show that GIFGuard achieves high-fidelity visual quality and remarkable robustness performance against deepfakes. Related code and dataset will be released.