# Harnessing Chain-of-Thought Reasoning in Multimodal Large Language Models for Face Anti-Spoofing

arXiv: https://arxiv.org/abs/2506.01783 · HTML: https://mlsanity.com/paper/2506.01783/

## Key facts

- Cited by 5 papers in the map (data through 14 Oct 2026).
- Placed in Multimodal reasoning, one of 17 directions in LLM reasoning.
- Builds on 37 of its 48 references that are in the map.
- First posted on arXiv 2 Jun 2025.

## Placed in

- LLM reasoning / [Multimodal reasoning](https://mlsanity.com/lr/multimodal-reasoning/)
- Multimodal LLMs / vision-language models / [MLLM jailbreaks and adversarial robustness](https://mlsanity.com/mllm/mllm-jailbreaks-and-adversarial-robustness/)

## Builds on (most cited first, 30 of 37)

- [Learning Transferable Visual Models From Natural Language Supervision](https://mlsanity.com/paper/2103.00020/) (2021)
- [GPT-4 Technical Report](https://mlsanity.com/paper/2303.08774/) (2023)
- [LoRA: Low-Rank Adaptation of Large Language Models](https://mlsanity.com/paper/2106.09685/) (2021)
- [Chain-of-Thought Prompting Elicits Reasoning in Large Language Models](https://mlsanity.com/paper/2201.11903/) (2022)
- [Self-Consistency Improves Chain of Thought Reasoning in Language Models](https://mlsanity.com/paper/2203.11171/) (2022)
- [Qwen2.5-VL Technical Report](https://mlsanity.com/paper/2502.13923/) (2025)
- [Expanding Performance Boundaries of Open-Source Multimodal Models with Model, Data, and Test-Time Scaling](https://mlsanity.com/paper/2412.05271/) (2024)
- [MiniCPM-V: A GPT-4V Level MLLM on Your Phone](https://mlsanity.com/paper/2408.01800/) (2024)
- [Visual-RFT: Visual Reinforcement Fine-Tuning](https://mlsanity.com/paper/2503.01785/) (2025)
- [LLaVA-CoT: Let Vision Language Models Reason Step-by-Step](https://mlsanity.com/paper/2411.10440/) (2024)
- [Learning Deep Models for Face Anti-Spoofing: Binary or Auxiliary Supervision](https://mlsanity.com/paper/1803.11097/) (2018)
- [CelebA-Spoof: Large-Scale Face Anti-Spoofing Dataset with Rich Annotations](https://mlsanity.com/paper/2007.12342/) (2020)
- [NAS-FAS: Static-Dynamic Central Difference Network Search for Face Anti-Spoofing](https://mlsanity.com/paper/2011.02062/) (2020)
- [face anti-spoofing based on color texture analysis](https://mlsanity.com/paper/1511.06316/) (2015)
- [Deep Tree Learning for Zero-shot Face Anti-Spoofing](https://mlsanity.com/paper/1904.02860/) (2019)
- [Domain Generalization via Shuffled Style Assembly for Face Anti-Spoofing](https://mlsanity.com/paper/2203.05340/) (2022)
- [PatchNet: A Simple Face Anti-Spoofing Framework via Fine-Grained Patch Recognition](https://mlsanity.com/paper/2203.14325/) (2022)
- [Rethinking Domain Generalization for Face Anti-spoofing: Separability and Alignment](https://mlsanity.com/paper/2303.13662/) (2023)
- [Revisiting Pixel-Wise Supervision for Face Anti-Spoofing](https://mlsanity.com/paper/2011.12032/) (2020)
- [FLIP: Cross-domain Face Anti-spoofing with Language Guidance](https://mlsanity.com/paper/2309.16649/) (2023)
- [Contrastive Context-Aware Learning for 3D High-Fidelity Mask Face Presentation Attack Detection](https://mlsanity.com/paper/2104.06148/) (2021)
- [Instance-Aware Domain Generalization for Face Anti-Spoofing](https://mlsanity.com/paper/2304.05640/) (2023)
- [Multi-domain Learning for Updating Face Anti-spoofing Models](https://mlsanity.com/paper/2208.11148/) (2022)
- [CFPL-FAS: Class Free Prompt Learning for Generalizable Face Anti-spoofing](https://mlsanity.com/paper/2403.14333/) (2024)
- [Adaptive Transformers for Robust Few-shot Cross-domain Face Anti-spoofing](https://mlsanity.com/paper/2203.12175/) (2022)
- [Chain of Thought Prompt Tuning in Vision Language Models](https://mlsanity.com/paper/2304.07919/) (2023)
- [Generative Domain Adaptation for Face Anti-Spoofing](https://mlsanity.com/paper/2207.10015/) (2022)
- [Unified Detection of Digital and Physical Face Attacks](https://mlsanity.com/paper/2104.02156/) (2021)
- [Gradient Alignment for Cross-Domain Face Anti-Spoofing](https://mlsanity.com/paper/2402.18817/) (2024)
- [Feature Generation and Hypothesis Verification for Reliable Face Anti-Spoofing](https://mlsanity.com/paper/2112.14894/) (2021)

## Built on by (most cited first, 5 of 5)

- [METER: Multi-modal Evidence-based Thinking and Explainable Reasoning -- Algorithm and Benchmark](https://mlsanity.com/paper/2507.16206/) (2025)
- [UniShield: Unified Face Attack Detection via KG-Informed Multimodal Reasoning](https://mlsanity.com/paper/2605.08709/) (2026)
- [FAS-R1: A Unified Multi-Task MLLM for Reasoning Face Anti-Spoofing](https://mlsanity.com/paper/2607.26432/) (2026)
- [LoRA-based Adaptation Alone Is Not Enough: Understanding the Limits of Foundation Models for Face Presentation Attack Detection](https://mlsanity.com/paper/2608.09633/) (2026)
- [Primitive-Driven Compositional Forensic Visual Prompting for Open-World Face Anti-Spoofing](https://mlsanity.com/paper/2608.17351/) (2026)

## Abstract (arXiv)

Face Anti-Spoofing (FAS) typically depends on a single visual modality when defending against presentation attacks such as print attacks, screen replays, and 3D masks, resulting in limited generalization across devices, environments, and attack types. Meanwhile, Multimodal Large Language Models (MLLMs) have recently achieved breakthroughs in image-text understanding and semantic reasoning, suggesting that integrating visual and linguistic co-inference into FAS can substantially improve both robustness and interpretability. However, the lack of a high-quality vision-language multimodal dataset has been a critical bottleneck. To address this, we introduce FaceCoT (Face Chain-of-Thought), the first large-scale Visual Question Answering (VQA) dataset tailored for FAS. FaceCoT covers 14 spoofing attack types and enriches model learning with high-quality CoT VQA annotations. Meanwhile, we develop a caption model refined via reinforcement learning to expand the dataset and enhance annotation quality. Furthermore, we introduce a CoT-Enhanced Progressive Learning (CEPL) strategy to better leverage the CoT data and boost model performance on FAS tasks. Extensive experiments demonstrate that models trained with FaceCoT and CEPL outperform state-of-the-art methods on multiple benchmark datasets.