HeadHunt-VAD: Hunting Robust Anomaly-Sensitive Heads in MLLM for Tuning-Free Video Anomaly Detection

• URL: http://arxiv.org/abs/2512.17601v2
• Date: Tue, 23 Dec 2025 10:13:38 GMT
• Title: HeadHunt-VAD: Hunting Robust Anomaly-Sensitive Heads in MLLM for Tuning-Free Video Anomaly Detection
• Authors: Zhaolin Cai, Fan Li, Ziwei Zheng, Haixia Bi, Lijun He,
• Abstract summary: Video Anomaly Detection (VAD) aims to locate events that deviate from normal patterns in videos.<n>Recent tuning-free methods based on Multimodal Large Language Models (MLLMs) offer a promising alternative by leveraging their rich world knowledge.<n>We propose HeadHunt-VAD, a novel tuning-free VAD paradigm that bypasses textual generation by directly hunting robust anomaly-sensitive internal attention heads.
• Score: 9.217348688177298
• License: http://creativecommons.org/licenses/by-nc-nd/4.0/
• Abstract: Video Anomaly Detection (VAD) aims to locate events that deviate from normal patterns in videos. Traditional approaches often rely on extensive labeled data and incur high computational costs. Recent tuning-free methods based on Multimodal Large Language Models (MLLMs) offer a promising alternative by leveraging their rich world knowledge. However, these methods typically rely on textual outputs, which introduces information loss, exhibits normalcy bias, and suffers from prompt sensitivity, making them insufficient for capturing subtle anomalous cues. To address these constraints, we propose HeadHunt-VAD, a novel tuning-free VAD paradigm that bypasses textual generation by directly hunting robust anomaly-sensitive internal attention heads within the frozen MLLM. Central to our method is a Robust Head Identification module that systematically evaluates all attention heads using a multi-criteria analysis of saliency and stability, identifying a sparse subset of heads that are consistently discriminative across diverse prompts. Features from these expert heads are then fed into a lightweight anomaly scorer and a temporal locator, enabling efficient and accurate anomaly detection with interpretable outputs. Extensive experiments show that HeadHunt-VAD achieves state-of-the-art performance among tuning-free methods on two major VAD benchmarks while maintaining high efficiency, validating head-level probing in MLLMs as a powerful and practical solution for real-world anomaly detection.

Related papers

• Steering and Rectifying Latent Representation Manifolds in Frozen Multi-modal LLMs for Video Anomaly Detection [52.5174167737992]
  Video anomaly detection (VAD) aims to identify abnormal events in videos.<n>We propose SteerVAD, which advances MLLM-based VAD by shifting from passively reading to actively steering and rectifying internal representations.<n>Our method achieves state-of-the-art performance among tuning-free approaches requiring only 1% of training data.
  arXiv  Detail & Related papers  (2026-02-27T13:48:50Z)
• HiProbe-VAD: Video Anomaly Detection via Hidden States Probing in Tuning-Free Multimodal LLMs [8.18063726177317]
  Video Anomaly Detection (VAD) aims to identify and locate deviations from normal patterns in video sequences.<n>We propose HiProbe-VAD, a novel framework that leverages pre-trained Multimodal Large Language Models (MLLMs) for VAD without requiring fine-tuning.
  arXiv  Detail & Related papers  (2025-07-23T10:41:46Z)
• Uncertainty-Aware Attention Heads: Efficient Unsupervised Uncertainty Quantification for LLMs [129.79394562739705]
  Large language models (LLMs) exhibit impressive fluency, but often produce critical errors known as "hallucinations"<n>We propose RAUQ (Recurrent Attention-based Uncertainty Quantification), an unsupervised approach that leverages intrinsic attention patterns in transformers to detect hallucinations efficiently.<n> Experiments across 4 LLMs and 12 question answering, summarization, and translation tasks demonstrate that RAUQ yields excellent results.
  arXiv  Detail & Related papers  (2025-05-26T14:28:37Z)
• SlowFastVAD: Video Anomaly Detection via Integrating Simple Detector and RAG-Enhanced Vision-Language Model [52.47816604709358]
  Video anomaly detection (VAD) aims to identify unexpected events in videos and has wide applications in safety-critical domains.<n> vision-language models (VLMs) have demonstrated strong multimodal reasoning capabilities, offering new opportunities for anomaly detection.<n>We propose SlowFastVAD, a hybrid framework that integrates a fast anomaly detector with a slow anomaly detector.
  arXiv  Detail & Related papers  (2025-04-14T15:30:03Z)
• VMAD: Visual-enhanced Multimodal Large Language Model for Zero-Shot Anomaly Detection [19.79027968793026]
  Zero-shot anomaly detection (ZSAD) recognizes and localizes anomalies in previously unseen objects. Existing ZSAD methods are limited by closed-world settings, struggling to unseen defects with predefined prompts. We propose a novel framework VMAD (Visual-enhanced MLLM Anomaly Detection) that enhances MLLM with visual-based IAD knowledge and fine-grained perception.
  arXiv  Detail & Related papers  (2024-09-30T09:51:29Z)
• Appearance Blur-driven AutoEncoder and Motion-guided Memory Module for Video Anomaly Detection [14.315287192621662]
  Video anomaly detection (VAD) often learns the distribution of normal samples and detects the anomaly through measuring significant deviations. Most VADs cannot cope with cross-dataset validation for new target domains. We propose a novel VAD method with a motion-guided memory module to achieve cross-dataset validation with zero-shot.
  arXiv  Detail & Related papers  (2024-09-26T07:48:20Z)
• Holmes-VAD: Towards Unbiased and Explainable Video Anomaly Detection via Multi-modal LLM [35.06386971859359]
  Holmes-VAD is a novel framework that leverages precise temporal supervision and rich multimodal instructions. We construct the first large-scale multimodal VAD instruction-tuning benchmark, VAD-Instruct50k. Building upon the VAD-Instruct50k dataset, we develop a customized solution for interpretable video anomaly detection.
  arXiv  Detail & Related papers  (2024-06-18T03:19:24Z)
• A Comprehensive Library for Benchmarking Multi-class Visual Anomaly Detection [89.92916473403108]
  This paper proposes a comprehensive visual anomaly detection benchmark, ADer, which is a modular framework for new methods.<n>The benchmark includes multiple datasets from industrial and medical domains, implementing fifteen state-of-the-art methods and nine comprehensive metrics.<n>We objectively reveal the strengths and weaknesses of different methods and provide insights into the challenges and future directions of multi-class visual anomaly detection.
  arXiv  Detail & Related papers  (2024-06-05T13:40:07Z)
• Towards Open Set Video Anomaly Detection [11.944167192592905]
  Open Set Video Anomaly Detection (OpenVAD) aims to identify abnormal events from video data where both known anomalies and novel ones exist in testing. We develop a novel weakly supervised method for the OpenVAD problem by integrating evidential deep learning (EDL) and normalizing flows (NFs) into a multiple instance learning (MIL) framework.
  arXiv  Detail & Related papers  (2022-08-23T17:53:34Z)
• Self-trained Deep Ordinal Regression for End-to-End Video Anomaly Detection [114.9714355807607]
  We show that applying self-trained deep ordinal regression to video anomaly detection overcomes two key limitations of existing methods. We devise an end-to-end trainable video anomaly detection approach that enables joint representation learning and anomaly scoring without manually labeled normal/abnormal data.
  arXiv  Detail & Related papers  (2020-03-15T08:44:55Z)
• SUOD: Accelerating Large-Scale Unsupervised Heterogeneous Outlier Detection [63.253850875265115]
  Outlier detection (OD) is a key machine learning (ML) task for identifying abnormal objects from general samples. We propose a modular acceleration system, called SUOD, to address it.
  arXiv  Detail & Related papers  (2020-03-11T00:22:50Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.