AI 中文总结
本文提出面向推理的两阶段多任务MLLM框架FAS-R1，结合DSA与DA-GRPO优化，在人脸防伪的分类、识别、定位任务中表现优异，性能优于对比系统且具备良好缩放性。
AI 中文摘要
人脸防伪（FAS）不仅需要做出真实/伪造判断，还需提供攻击语义和图像层面证据供人工检查。现有判别式FAS模型多以标签为中心，而近期基于多模态大语言模型（MLLM）的方法虽能生成结构化输出，但仍主要依赖监督微调，常产生模板化理由且对难攻击的优化不足。本文提出FAS-R1，这是一种面向推理的两阶段统一FAS预测MLLM框架，涵盖真实性分类、攻击类型识别和伪造区域定位。FAS-R1首先使用高质量长思维链数据集FAS-R1-23K进行冷启动监督微调，随后执行FAS专用的广义偏好优化（GRPO）后训练。退化模拟增强（DSA）可鼓励模型在视觉质量变化下稳定推理伪造线索，而难度感知GRPO（DA-GRPO）可缓解易样本主导问题，避免难任务-攻击组（尤其是化妆、面具等微妙或模糊攻击）优化不足。3B规模的FAS-R1主模型在域内实现98.75%的真实性准确率、93.33%的攻击类型准确率，以及96.30/94.73%的AP@40/AP@50指标，在跨域真实性泛化和答案与理由质量上均优于对比系统，不同基础模型的实验也显示出良好的缩放特性，代码即将发布。
英文摘要
Face anti-spoofing (FAS) is increasingly expected to provide not only bona fide/spoof decisions, but also attack semantics and image-grounded evidence for human inspection. Existing discriminative FAS models remain largely label-centric, while recent MLLM-based methods offer structured outputs but still rely mainly on supervised fine-tuning, often producing template-like rationales and weak optimization for difficult attacks. We propose FAS-R1, a two-stage reasoning-oriented MLLM framework for unified FAS prediction, covering authenticity classification, attack-type recognition and spoof-region localization. FAS-R1 first uses FAS-R1-23K, a high-quality long-CoT dataset, for cold-start supervised fine-tuning, and then performs FAS-specific GRPO post-training. Degradation-Simulated Augmentation (DSA) encourages stable spoof-cue reasoning across visual-quality shifts, while Difficulty-Aware GRPO (DA-GRPO) mitigates easy-sample dominance that may leave difficult task--attack groups under-optimized, especially for subtle or ambiguous attacks such as makeup and mask attacks. The main 3B FAS-R1 model achieves 98.75\% authenticity accuracy, 93.33\% attack-type accuracy, and 96.30/94.73\% AP@40/AP@50 in-domain. It also outperforms the compared systems in cross-domain authenticity generalization and answer-and-rationale quality. Experiments with different base models further show favorable scaling behavior. The code will be released soon.
