Logo image
Back
Seeing the Unseen: Enhancing Synthetic Video Detector Transferability to New AI Video Generation Models via Virtual Generators
Conference proceeding   Open access

Seeing the Unseen: Enhancing Synthetic Video Detector Transferability to New AI Video Generation Models via Virtual Generators

Danial Samadi Vahdati, Tai Duc Nguyen, Aref Azizpour, Hamed Ahangari and Matthew Stamm
Proceedings of the 2026 ACM Workshop on Information Hiding and Multimedia Security, pp 241-252
17 Jun 2026
DOI: https://doi.org/10.1145/3785353.3815083
Featured in Collection :   Research Supported by Drexel Libraries' OA Programs
url
https://doi.org/10.1145/3785353.3815083View
Published, Version of Record (VoR) Open Access via Drexel Libraries Read and Publish Program 2026 Open CC BY-NC-ND V4.0

Abstract

Computing methodologies -- Image processing
Synthetic video detectors often struggle to generalize to content generated by previously unseen models. While training on a diverse set of generators can improve transferability, this approach is inherently limited by the finite number of known generators. In this work, we address this challenge by introducing a novel technique to artificially boost training diversity through virtual generators - parametric models that synthesize forensic microstructures not associated with any real generator. We model these microstructures using a 2D autoregressive process and design AR parameters that simulate plausible yet unseen generative behaviors, informed by architectural patterns and learned transformations in modern video generators. Our results demonstrate that detectors trained with virtual generators significantly outperform those trained with traditional data augmentation and offer improved generalization to new generators.

Metrics

1 Record Views

Details

Logo image