From Intuition to Investigation: A Tool-Augmented Reasoning MLLM Framework for Generalizable Face Anti-Spoofing

2026-03-01

AI Summary

This work addresses the limited cross-domain generalization of existing multimodal large language model (MLLM)-based face anti-spoofing (FAS) methods, which rely on coarse-grained semantic descriptions and struggle to capture fine-grained forgery cues. To overcome this, we propose TAR-FAS, a novel framework that reformulates FAS as a chain-of-thought reasoning process augmented with visual tools, enabling the model to start from intuitive observations and adaptively invoke external tools for in-depth analysis of subtle spoofing artifacts. We introduce the first tool-augmented multi-turn reasoning paradigm for FAS, construct ToolFAS-16K—a dataset containing tool usage trajectories—and design a Diverse-Tool Group Relative Policy Optimization (DT-GRPO) algorithm to enable efficient and autonomous invocation of diverse visual tools. Under the challenging 1-vs-11 cross-domain protocol, TAR-FAS achieves state-of-the-art performance, significantly improving both generalization and interpretability.

Technology Category

• Machine Learning: Large Multimodal Models (LMMs)
• Computer Vision: Multi-modal Vision
• Natural Language Processing: Fact-Checking / Misinformation Detection (NLP Focus)

Application Category

• Search and Retrieval-Augmented AI: Search Tool Learning with LLM: Teaching LLMs to invoke search and make use of retrieved information
• Responsible Web: Machine-in-the-loop, human agency and autonomy
• Systems and Infrastructure for Web, Mobile and WoT: Applied ML and AI for Web-based mobile applications

Abstract

Face recognition remains vulnerable to presentation attacks, calling for robust Face Anti-Spoofing (FAS) solutions. Recent MLLM-based FAS methods reformulate the binary classification task as the generation of brief textual descriptions to improve cross-domain generalization. However, their generalizability is still limited, as such descriptions mainly capture intuitive semantic cues (e.g., mask contours) while struggling to perceive fine-grained visual patterns. To address this limitation, we incorporate external visual tools into MLLMs to encourage deeper investigation of subtle spoof clues. Specifically, we propose the Tool-Augmented Reasoning FAS (TAR-FAS) framework, which reformulates the FAS task as a Chain-of-Thought with Visual Tools (CoT-VT) paradigm, allowing MLLMs to begin with intuitive observations and adaptively invoke external visual tools for fine-grained investigation. To this end, we design a tool-augmented data annotation pipeline and construct the ToolFAS-16K dataset, which contains multi-turn tool-use reasoning trajectories. Furthermore, we introduce a tool-aware FAS training pipeline, where Diverse-Tool Group Relative Policy Optimization (DT-GRPO) enables the model to autonomously learn efficient tool use. Extensive experiments under a challenging one-to-eleven cross-domain protocol demonstrate that TAR-FAS achieves SOTA performance while providing fine-grained visual investigation for trustworthy spoof detection.

Problem

• Face Anti-Spoofing
• cross-domain generalization
• fine-grained visual patterns
• presentation attacks
• MLLM

Innovation

• Tool-Augmented Reasoning
• Chain-of-Thought with Visual Tools
• Face Anti-Spoofing
• Cross-Domain Generalization
• Multimodal Large Language Model