Search papers, labs, and topics across Lattice.
This paper benchmarks 15 vision-only foundation models for face anti-spoofing (FAS) under challenging cross-domain scenarios. They find that self-supervised vision models, specifically DINOv2 with Registers, excel at capturing fine-grained spoofing cues while suppressing attention artifacts. By combining DINOv2 with FAS-Aug, PDA, and APL, the proposed vision-only baseline achieves state-of-the-art performance on the MICO protocol and outperforms existing methods on the LSD protocol with superior computational efficiency.
Forget multimodal bloat: optimized vision-only transformers can achieve state-of-the-art face anti-spoofing, outperforming complex VLMs with far greater efficiency.
Face Anti-Spoofing (FAS) remains challenging due to the requirement for robust domain generalization across unseen environments. While recent trends leverage Vision-Language Models (VLMs) for semantic supervision, these multimodal approaches often demand prohibitive computational resources and exhibit high inference latency. Furthermore, their efficacy is inherently limited by the quality of the underlying visual features. This paper revisits the potential of vision-only foundation models to establish a highly efficient and robust baseline for FAS. We conduct a systematic benchmarking of 15 pre-trained models, such as supervised CNNs, supervised ViTs, and self-supervised ViTs, under severe cross-domain scenarios including the MICO and Limited Source Domains (LSD) protocols. Our comprehensive analysis reveals that self-supervised vision models, particularly DINOv2 with Registers, significantly suppress attention artifacts and capture critical, fine-grained spoofing cues. Combined with Face Anti-Spoofing Data Augmentation (FAS-Aug), Patch-wise Data Augmentation (PDA) and Attention-weighted Patch Loss (APL), our proposed vision-only baseline achieves state-of-the-art performance in the MICO protocol. This baseline outperforms existing methods under the data-constrained LSD protocol while maintaining superior computational efficiency. This work provides a definitive vision-only baseline for FAS, demonstrating that optimized self-supervised vision transformers can serve as a backbone for both vision-only and future multimodal FAS systems. The project page is available at: https://gsisaoki.github.io/FAS-VFMbenchmark-CVPRW2026/ .
