mlsanityData to 7 Oct 2026

Unlocking Vision-Language Models for Video Anomaly Detection via Fine-Grained Prompting

arXiv 2510.02155 ↗First posted 2 Oct 2025cs.CV cs.AIcited by 8

On the map: Multimodal LLMs / vision-language models → Chart understanding and anomaly detection

Abstract (arXiv)

Prompting has emerged as a practical way to adapt frozen vision-language models (VLMs) for video anomaly detection (VAD). Yet, existing prompts are often overly abstract, overlooking the fine-grained human-object interactions or action semantics that define complex anomalies in surveillance videos. We propose ASK-Hint, a structured prompting framework that leverages action-centric knowledge to elicit more accurate and interpretable reasoning from frozen VLMs. Our approach organizes prompts into semantically coherent groups (e.g. violence, property crimes, public safety) and formulates fine-grained guiding questions that align model predictions with discriminative visual cues. Extensive experiments on UCF-Crime and XD-Violence show that ASK-Hint consistently improves AUC over prior baselines, achieving state-of-the-art performance compared to both fine-tuned and training-free methods. Beyond accuracy, our framework provides interpretable reasoning traces towards anomaly and demonstrates strong generalization across datasets and VLM backbones. These results highlight the critical role of prompt granularity and establish ASK-Hint as a new training-free and generalizable solution for explainable video anomaly detection.

Key facts. Cited by 8 papers in the map (data through 7 Oct 2026). Placed in Chart understanding and anomaly detection, one of 16 directions in Multimodal LLMs / vision-language models. Builds on 44 of its 55 references that are in the map. First posted on arXiv 2 Oct 2025.

Builds on

44 of 55 references.

Closest ancestors · 2

Same direction, newest first

  1. 2024Holmes-VAD: Towards Unbiased and Explainable Video Anomaly Detection via Multi-modal LLM18 Jun
  2. 2023VadCLIP: Adapting Vision-Language Models for Weakly Supervised Video Anomaly Detection22 Aug

Other references · 42

  1. 2023GPT-4 Technical Reportcited by 18,132
  2. 2025Qwen2.5-VL Technical Reportcited by 5,418
  3. 2023Improved Baselines with Visual Instruction Tuningcited by 4,362
  4. 2023InternVL: Scaling up Vision Foundation Models and Aligning for Generic Visual-Linguistic Taskscited by 2,534
  5. 2023Video-LLaVA: Learning United Visual Representation by Alignment Before Projectioncited by 1,385
  6. 2018Real-world Anomaly Detection in Surveillance Videoscited by 570
  7. 2025Video-R1: Reinforcing Video Reasoning in MLLMscited by 464
  8. 2017Future Frame Prediction for Anomaly Detection -- A New Baselinecited by 391
  9. 2024InternVideo2: Scaling Foundation Models for Multimodal Video Understandingcited by 310
  10. 2016Learning Temporal Regularity in Video Sequencescited by 254
  11. 2020Not only Look, but also Listen: Learning Multimodal Violence Detection under Weak Supervisioncited by 235
  12. 2021Weakly-supervised Video Anomaly Detection with Robust Temporal Feature Magnitude Learningcited by 226
30 more
  1. 2021MIST: Multiple Instance Self-Training Framework for Video Anomaly Detectioncited by 137
  2. 2022PromptFL: Let Federated Participants Cooperatively Learn Prompts Instead of Models -- Federated Learning in Age of Foundation Modelcited by 133
  3. 2022Generative Cooperative Learning for Unsupervised Video Anomaly Detectioncited by 102
  4. 2023Dual Memory Units with Uncertainty Regulation for Weakly Supervised Video Anomaly Detectioncited by 88
  5. 2022MGFN: Magnitude-Contrastive Glance-and-Focus Network for Weakly-Supervised Video Anomaly Detectioncited by 87
  6. 2024Harnessing Large Language Models for Training-free Video Anomaly Detectioncited by 85
  7. 2022CLIP-TSA: CLIP-Assisted Temporal Self-Attention for Weakly-Supervised Video Anomaly Detectioncited by 67
  8. 2019GODS: Generalized One-class Discriminative Subspaces for Anomaly Detectioncited by 64
  9. 2023Generalized Video Anomaly Event Detection: Systematic Taxonomy and Comparison of Deep Modelscited by 59
  10. 2023Open-Vocabulary Video Anomaly Detectioncited by 57
  11. 2024Follow the Rules: Reasoning for Video Anomaly Detection with Large Language Modelscited by 55
  12. 2024Text Prompt with Normality Guidance for Weakly Supervised Video Anomaly Detectioncited by 49
  13. 2024Video Anomaly Detection and Explanation via Large Language Modelscited by 45
  14. 2023ArGue: Attribute-Guided Prompt Tuning for Vision-Language Modelscited by 39
  15. 2025EventVAD: Training-Free Event-Aware Video Anomaly Detectioncited by 32
  16. 2022DyAnNet: A Scene Dynamicity Guided Self-Trained Video Anomaly Detection Networkcited by 26
  17. 2025More Thought, Less Accuracy? On the Dual Nature of Reasoning in Vision-Language Modelscited by 26
  18. 2025Identifying and Mitigating Position Bias of Multi-image Vision-Language Modelscited by 24
  19. 2025VAU-R1: Advancing Video Anomaly Understanding via Reinforcement Fine-Tuningcited by 24
  20. 2025MASH-VLM: Mitigating Action-Scene Hallucination in Video-LLMs through Disentangled Spatial-Temporal Representationscited by 22
  21. 2019Anomaly Locality in Video Surveillancecited by 18
  22. 2023Unsupervised Video Anomaly Detection with Diffusion Models Conditioned on Compact Motion Representationscited by 18
  23. 2024Collaborative Learning of Anomalies with Privacy (CLAP) for Unsupervised Video Anomaly Detection: A New Baselinecited by 17
  24. 2025Flashback: Memory-Driven Zero-shot, Real-time Video Anomaly Detectioncited by 12
  25. 2025Temporal Alignment-Free Video Matching for Few-shot Action Recognitioncited by 10
  26. 2025HiProbe-VAD: Video Anomaly Detection via Hidden States Probing in Tuning-Free Multimodal LLMscited by 8
  27. 2023Exploring the Potential of World Models for Anomaly Detection in Autonomous Drivingcited by 7
  28. 2025Black Sheep in the Herd: Playing with Spuriously Correlated Attributes for Vision-Language Recognitioncited by 7
  29. 2025SimLabel: Consistency-Guided OOD Detection with Pretrained Vision-Language Modelscited by 5
  30. 2022Spatio-Temporal Relation Learning for Video Anomaly Detectioncited by 1

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2025: 1; 2026: 7.

8 papers.

  1. 2026Understanding Multimodal Complementarity for Single-Frame Action Anticipationcited by 2
  2. 2025High-Entropy Tokens as Multimodal Failure Points in Vision-Language Modelscited by 1
  3. 2026Reasoning-Guided Grounding: Elevating Video Anomaly Detection through Multimodal Large Language Modelscited by 1
  4. 2026COPRA: Conditional Parameter Adaptation with Reinforcement Learning for Video Anomaly Detectioncited by 1
  5. 2026Break the Brake, Not the Wheel: Untargeted Jailbreak via Entropy Maximizationcited by 0
  6. 2026O-VAD: Industrial Video Anomaly Detection through Object-Centric Tracking and Reasoningcited by 0
  7. 2026NOVA: Normal-Side Modeling for Training-Free Zero-Shot Video Anomaly Detectioncited by 0
  8. 2026AnomalyCraft-700K: Component-Level Controllable and Verifiable Synthetic Anomalies for Fine-Grained Video Anomaly Understandingcited by 0

Similar, not cited

Close in wording, no citation either way

  1. 2026QVAD: A Question-Centric Agentic Framework for Efficient and Training-Free Video Anomaly Detection
  2. 2026CoReVAD: A Contextual Reasoning Framework for Training-Free Video Anomaly Detection
  3. 2024VERA: Explainable Video Anomaly Detection via Verbalized Learning of Vision-Language Models
  4. 2026PrismVAU: Prompt-Refined Inference System for Multimodal Video Anomaly Understanding
  5. 2026Glance, Scrutinize, and Think: Advancing Video Anomaly Detection from Training-Free to Agentic Reasoning
  6. 2025Knowledge-Guided Textual Reasoning for Explainable Video Anomaly Detection via LLMs
  7. 2026LATERN: Test-Time Context-Aware Explainable Video Anomaly Detection
  8. 2026GridVAD: Open-Set Video Anomaly Detection via Spatial Reasoning over Stratified Frame Grids