mlsanityData to 14 Oct 2026

CPG-PAD: Concept-Informed Prompts Guided Presentation Attack Detection

arXiv 2607.01303 ↗First posted 1 Jul 2026cs.CV cs.AIcited by 1

Not on the map. Why?

Abstract (arXiv)

Presentation Attack Detection (PAD) serves as a crucial safeguard for face recognition systems against presentation attacks such as printed photos, replayed videos, and 3D masks. Despite significant progress, existing PAD models still struggle to generalize across unseen domains due to variations in sensors, lighting, and attack materials. Recent Vision-Language Models (VLMs) have shown strong generalization ability, yet their applications in PAD remain limited because learned prompts, typically optimized under class-label supervision, fail to explicitly align with fine-grained attack-relevant visual semantics. As a result, the learned representations often overfit domain-specific artifacts instead of capturing transferable attack cues. To address this, we propose Concept-Informed Prompts Guided Presentation Attack Detection (CPG-PAD), a framework that introduces model-level concept guidance into the prompt learning process. Specifically, we design a Visual Concept-driven Enhancement (VCE) module that employs eXplainable AI (XAI) techniques to automatically discover PAD-relevant visual concepts and generate concept-associated heatmaps providing localized fine-grained guidance. Guided by these heatmaps, a Prompt-based Concept Injection (PCI) mechanism integrates these concepts into the prompt space through a Visual-Prompt Decoder (VPD) and a concept-mapping loss, enabling prompts to align with the model's internal concept space. This design enables CPG-PAD to capture generalizable and domain-invariant attack cues while effectively suppressing dataset-specific biases. Extensive experiments across nine benchmark datasets demonstrate that CPG-PAD consistently achieves state-of-the-art cross-domain performance under multi-source, limited-source, and single-source settings.

Key facts. Cited by 1 paper in the map (data through 14 Oct 2026). Builds on 46 of its 83 references that are in the map. First posted on arXiv 1 Jul 2026.

Builds on

46 of 83 references.

References · 46

  1. 2017Attention Is All You Needcited by 60,361
  2. 2021Learning Transferable Visual Models From Natural Language Supervisioncited by 33,744
  3. 2016Grad-CAM: Visual Explanations from Deep Networks via Gradient-based Localizationcited by 6,733
  4. 2021Learning to Prompt for Vision-Language Modelscited by 2,281
  5. 2022Conditional Prompt Learning for Vision-Language Modelscited by 1,433
  6. 2017Interpretability Beyond Feature Attribution: Quantitative Testing with Concept Activation Vectors (TCAV)cited by 1,319
  7. 2017Adversarial Discriminative Domain Adaptationcited by 1,311
  8. 2017Grad-CAM++: Improved Visual Explanations for Deep Convolutional Networkscited by 939
  9. 2018RISE: Randomized Input Sampling for Explanation of Black-box Modelscited by 838
  10. 2016Synthesizing the preferred inputs for neurons in neural networks via deep generator networkscited by 181
  11. 2022CRAFT: Concept Recursive Activation FacTorization for Explainabilitycited by 178
  12. 2020Searching Central Difference Convolutional Networks for Face Anti-Spoofingcited by 173
34 more
  1. 2018Learning Deep Models for Face Anti-Spoofing: Binary or Auxiliary Supervisioncited by 167
  2. 2016Deep Reconstruction-Classification Networks for Unsupervised Domain Adaptationcited by 163
  3. 2020Single-Side Domain Generalization for Face Anti-Spoofingcited by 139
  4. 2020CelebA-Spoof: Large-Scale Face Anti-Spoofing Dataset with Rich Annotationscited by 109
  5. 2023A Holistic Approach to Unifying Automatic Concept Extraction and Concept Importance Estimationcited by 109
  6. 2019Regularized Fine-grained Meta Face Anti-spoofingcited by 86
  7. 2019Biometric Face Presentation Attack Detection with Multi-Channel Convolutional Neural Networkcited by 79
  8. 2022Domain Generalization via Shuffled Style Assembly for Face Anti-Spoofingcited by 73
  9. 2020Deep Spatial Gradient and Temporal Depth Learning for Face Anti-spoofingcited by 70
  10. 2020CASIA-SURF CeFA: A Benchmark for Multi-modal Cross-ethnicity Face Anti-spoofingcited by 66
  11. 2019CASIA-SURF: A Large-scale Multi-modal Benchmark for Face Anti-spoofingcited by 61
  12. 2021Self-Domain Adaptation for Face Anti-Spoofingcited by 49
  13. 2023Rethinking Domain Generalization for Face Anti-spoofing: Separability and Alignmentcited by 48
  14. 2021Dual Reweighting Domain Generalization for Face Presentation Attack Detectioncited by 47
  15. 2023FLIP: Cross-domain Face Anti-spoofing with Language Guidancecited by 45
  16. 2020On the Effectiveness of Vision Transformers for Zero-shot Face Anti-Spoofingcited by 44
  17. 2021Contrastive Context-Aware Learning for 3D High-Fidelity Mask Face Presentation Attack Detectioncited by 43
  18. 2023Instance-Aware Domain Generalization for Face Anti-Spoofingcited by 42
  19. 2022Multi-domain Learning for Updating Face Anti-spoofing Modelscited by 39
  20. 2024CFPL-FAS: Class Free Prompt Learning for Generalizable Face Anti-spoofingcited by 39
  21. 2022Adaptive Transformers for Robust Few-shot Cross-domain Face Anti-spoofingcited by 38
  22. 2022Generative Domain Adaptation for Face Anti-Spoofingcited by 33
  23. 2021Learning Meta Pattern for Face Anti-Spoofingcited by 29
  24. 2024Gradient Alignment for Cross-Domain Face Anti-Spoofingcited by 27
  25. 2021Feature Generation and Hypothesis Verification for Reliable Face Anti-Spoofingcited by 23
  26. 2024Test-Time Domain Generalization for Face Anti-Spoofingcited by 20
  27. 2025FoundPAD: Foundation Models Reloaded for Face Presentation Attack Detectioncited by 14
  28. 2025Interpretable Face Anti-Spoofing: Enhancing Generalization with Multimodal Large Language Modelscited by 11
  29. 2018Exploiting temporal and depth information for multi-frame face anti-spoofingcited by 10
  30. 2022Cyclically Disentangled Feature Translation for Face Anti-spoofingcited by 10
  31. 2025Domain Generalization for Face Anti-spoofing via Content-aware Composite Prompt Engineeringcited by 7
  32. 2022Learning Multiple Explainable and Generalizable Cues for Face Anti-spoofingcited by 5
  33. 2023Domain-Generalized Face Anti-Spoofing with Unknown Attackscited by 4
  34. 2026From Intuition to Investigation: A Tool-Augmented Reasoning MLLM Framework for Generalizable Face Anti-Spoofingcited by 2

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2026: 1.

1 paper.

  1. 2026Primitive-Driven Compositional Forensic Visual Prompting for Open-World Face Anti-Spoofingcited by 0

Similar, not cited

Close in wording, no citation either way

  1. 2025Learning Knowledge-based Prompts for Robust 3D Mask Presentation Attack Detection
  2. 2025Exploring ChatGPT for Face Presentation Attack Detection in Zero and Few-Shot in-Context Learning
  3. 2025Steering Vision-Language Pre-trained Models for Incremental Face Presentation Attack Detection
  4. 2026Foundation and Multimodal Large Language Models for Face Presentation and Morph Attack Detection
  5. 2022One-Class Knowledge Distillation for Face Presentation Attack Detection
  6. 2022VLAD-VSA: Cross-Domain Face Presentation Attack Detection with Vocabulary Separation and Adaptation
  7. 2023TeG-DG: Textually Guided Domain Generalization for Face Anti-Spoofing
  8. 2026DE-FIVE: Detecting Malicious Image Prompts via Fourier Features and Image Vector Embeddings