mlsanityData to 14 Oct 2026

MS-MFAD : Multimodal large language models for Face Anti-spoofing Detection

arXiv 2608.17328 ↗First posted 18 Aug 2026cs.CVcited by 0

On the map: Multimodal LLMs / vision-language models → MLLM jailbreaks and adversarial robustness

Abstract (arXiv)

Facial biometric recognition systems currently face compound threats intertwining generative AI and high-fidelity physical spoofing. Existing defenses suffer from systemic bottlenecks, including poor generalization, non-auditable reasoning, and reliance on massive, low-quality datasets. To address these challenges, we propose Multimodal Large Language Models (MFAD) for face anti-spoofing detection, an explainable reasoning system for Unified Face Anti-Spoofing Detection (UFAD), accompanied by a semantic-level annotation benchmark. Unlike methods relying on external tools or coarse alignment, MFAD activates the intrinsic reasoning capabilities of Multimodal Large Language Models (MLLMs) via a fine-grained pixel-semantic anchoring mechanism. This eliminates localization hallucinations and ensures auditable reasoning paths. We introduce a cross-attack semantic-level unified annotation paradigm: by annotating only 1,000 precise masks per attack category, we generate reasoning evidence chains strictly corresponding to spoofed regions. Supervised fine-tuning on the Qwen-VL foundation model demonstrates that, using limited high-quality samples, the system achieves a 40-50% relative reduction in in-domain ACER and restricts cross-domain performance degradation to within 11.62%/5.23%, significantly outperforming existing frameworks. Furthermore, under white-box adversarial attacks, detection accuracy drops by only 3.2%, validating the robustness of semantic anchoring compared to models trained on massive short-text data. Domain practitioners rated the evidence reliability of reasoning paths at 4.57/5, with inference latency satisfying real-time deployment requirements. These results confirm that a few-shot, high-quality semantic annotation paradigm is effective for building trustworthy, explainable, and cost-efficient UFAD systems.

Key facts. No citing papers in the map yet (data through 14 Oct 2026). Placed in MLLM jailbreaks and adversarial robustness, one of 16 directions in Multimodal LLMs / vision-language models. Builds on 29 of its 52 references that are in the map. First posted on arXiv 18 Aug 2026.

Builds on

29 of 52 references.

Closest ancestors · 2

Same direction, newest first

  1. 2025FaceShield: Explainable Face Anti-Spoofing with Multimodal Large Language Models14 May
  2. 2024FFAA: Multimodal Large Language Model based Explainable Open-World Face Forgery Analysis Assistant19 Aug

Other references · 27

  1. 2022Training language models to follow instructions with human feedbackcited by 15,971
  2. 2021LoRA: Low-Rank Adaptation of Large Language Modelscited by 14,380
  3. 2024DeepSeekMath: Pushing the Limits of Mathematical Reasoning in Open Language Modelscited by 9,379
  4. 2017Towards Deep Learning Models Resistant to Adversarial Attackscited by 6,415
  5. 2023Qwen-VL: A Versatile Vision-Language Model for Understanding, Localization, Text Reading, and Beyondcited by 2,224
  6. 2019FaceForensics++: Learning to Detect Manipulated Facial Imagescited by 1,101
  7. 2019Celeb-DF: A Large-scale Challenging Dataset for DeepFake Forensicscited by 575
  8. 2019The Deepfake Detection Challenge (DFDC) Preview Datasetcited by 209
  9. 2020Searching Central Difference Convolutional Networks for Face Anti-Spoofingcited by 173
  10. 2018Learning Deep Models for Face Anti-Spoofing: Binary or Auxiliary Supervisioncited by 167
  11. 2023Representational Strengths and Limitations of Transformerscited by 133
  12. 2020CelebA-Spoof: Large-Scale Face Anti-Spoofing Dataset with Rich Annotationscited by 109
15 more
  1. 2019Biometric Face Presentation Attack Detection with Multi-Channel Convolutional Neural Networkcited by 79
  2. 2019CASIA-SURF: A Large-scale Multi-modal Benchmark for Face Anti-spoofingcited by 61
  3. 2023Towards General Visual-Linguistic Face Forgery Detectioncited by 45
  4. 2023FLIP: Cross-domain Face Anti-spoofing with Language Guidancecited by 45
  5. 2024FaceXFormer: A Unified Transformer for Facial Analysiscited by 34
  6. 2021Unified Detection of Digital and Physical Face Attackscited by 31
  7. 20213D High-Fidelity Mask Face Presentation Attack Detection Challengecited by 19
  8. 2024Unified Physical-Digital Face Attack Detectioncited by 17
  9. 2025MLLM-Enhanced Face Forgery Detection: A Vision-Language Fusion Solutioncited by 15
  10. 2024Joint Physical-Digital Facial Attack Detection Via Simulating Spoofing Cluescited by 8
  11. 2024La-SoftMoE CLIP for Unified Physical-Digital Face Attack Detectioncited by 8
  12. 2025MFFI: Multi-Dimensional Face Forgery Image Dataset for Real-World Scenarioscited by 3
  13. 2024AttackNet: Enhancing Biometric Security via Tailored Convolutional Neural Network Architectures for Liveness Detectioncited by 2
  14. 2024A visualization method for data domain changes in CNN networks and the optimization method for selecting thresholds in classification taskscited by 2
  15. 2026From Intuition to Investigation: A Tool-Augmented Reasoning MLLM Framework for Generalizable Face Anti-Spoofingcited by 2

Built on by

Not cited yet.

Similar, not cited

Close in wording, no citation either way

  1. 2025Interpretable Face Anti-Spoofing: Enhancing Generalization with Multimodal Large Language Models
  2. 2025Harnessing Chain-of-Thought Reasoning in Multimodal Large Language Models for Face Anti-Spoofing
  3. 2024SHIELD : An Evaluation Benchmark for Face Spoofing and Forgery Detection with Multimodal Large Language Models
  4. 2026FAS-R1: A Unified Multi-Task MLLM for Reasoning Face Anti-Spoofing
  5. 2026Benchmarking Vision Foundation Models for Domain-Generalizable Face Anti-Spoofing
  6. 2025PA-FAS: Towards Interpretable and Generalizable Multimodal Face Anti-Spoofing via Path-Augmented Reinforcement Learning
  7. 2026Foundation and Multimodal Large Language Models for Face Presentation and Morph Attack Detection
  8. 2026UniShield: Unified Face Attack Detection via KG-Informed Multimodal Reasoning