Research map / Vision-language and multimodal / Cross-modal learning
Semantic Freedom Bottleneck for Domain-Generalized Multimodal Face Anti-Spoofing
Yingjie Ma, Haonan Wang, Xun Lin, Hui Ma, Ruixin Zhang, Jingyun Zhang, Jun Wang, Rizen Guo, Shouhong Ding, Weicheng Xie, Linlin Shen, Zitong YU
Read the paperCheck your idea against 137,217 papers
AI-extracted summary. Written by AI from the paper's abstract only. Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Revisits multimodal domain-generalized face anti-spoofing from a representation-freedom view, using a CLIP-based post-fusion baseline to study how encoder depth affects cross-domain generalization.
Problem
Multimodal face anti-spoofing (RGB, infrared, depth) generalizes poorly to unseen domains; increasing visual encoder depth does not monotonically improve cross-domain performance.
Approach
Frames multimodal domain-generalized FAS through representation freedom, arguing that unconstrained semantic freedom in deep encoders hurts generalization and proposing a semantic freedom bottleneck to constrain it. Based on the abstract only; details of the bottleneck are not stated.
Surprising findings
- Increasing visual encoder depth in a CLIP-based post-fusion baseline does not monotonically improve cross-domain performance.
Experimental setup
Builds on: CLIP
Closest papers by meaning
- Interpretable Face Anti-Spoofing: Enhancing Generalization with Multimodal Large Language Models (AAAI 2025 · similarity 0.90)
- mmFAS: Multimodal Face Anti-Spoofing Using Multi-Level Alignment and Switch-Attention Fusion (AAAI 2025 · similarity 0.90)
- FaceShield: Explainable Face Anti-Spoofing with Multimodal Large Language Models (AAAI 2026 · similarity 0.90)
- From Intuition to Investigation: A Tool-Augmented Reasoning MLLM Framework for Generalizable Face Anti-Spoofing (CVPR 2026 · similarity 0.89)
- SLIP: Spoof-Aware One-Class Face Anti-Spoofing with Language Image Pretraining (AAAI 2025 · similarity 0.88)
- PA-FAS: Towards Interpretable and Generalizable Multimodal Face Anti-Spoofing via Path-Augmented Reinforcement Learning (AAAI 2026 · similarity 0.88)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
