Multimodal Anomaly Detection

Follow this topic Ask about this topic

The field studies how to detect and locate events or objects that differ from normal behavior by combining signals that may reveal different aspects of an anomaly. Common approaches model normality through reconstruction or feature distributions, compare modalities for inconsistencies, and fuse image, depth, point-cloud, video, text, or sensor features; newer methods also adapt vision-language models for grounded detection and explanation. Applications range from industrial defect inspection to surveillance, medical imaging, robotics, and monitoring complex systems, with evaluation spanning image- and pixel-level benchmarks as well as time-series and video tasks. A central challenge is making fusion reliable when modalities are noisy, missing, or unevenly informative, while ensuring that language-guided reasoning and benchmark performance reflect genuine visual or sensor evidence and transfer across settings.

Latest papers

97 papers14 papers in the last four weeks, up 75% on the four weeks before. 0.1% of all new papers.

EAGLE: Expert-Augmented Attention Guidance for Tuning-Free Industrial Anomaly Detection in Multimodal Large Language Models
Feb 19, 2026Xiaomeng Peng, Xilang Huang, Seon Han ChoiAnomaly Detection Industrial Anomaly Detection
Ewha Womans University
Towards Explainable Industrial Anomaly Detection via Knowledge-Guided Latent Reasoning
Feb 10, 2026Peng Chen, Wenqiang Wang, Yunkang Cao +7Industrial Anomaly Detection Multimodal Anomaly Detection
School of Cyber Science and Technology, Shenzhen Campus of Sun Yat-sen University, Shenzhen, China · School of Artificial Intelligence and Robotics, Hunan University, Changsha, China · Department of Computer and Information Science, University of Macau, Macau, China
MedAD-R1: Consistency-Reinforced Policy Optimization for Interpretable Medical Anomaly Detection
Feb 1, 2026Haitao Zhang, Yingying Wang, Jiaxiang Wang +5Healthcare Group Relative Policy Optimization
School of Informatics, Xiamen University · Institute of Artificial Intelligence, Xiamen University · Zhejiang Expressway Co., Ltd. +3
Defect-aware Hybrid Prompt Optimization via Progressive Tuning for Zero-Shot Multi-type Anomaly Detection and Segmentation
Dec 10, 2025Nadeem Nazer, Hongkuan Zhou, Lavdim Halilaj +2Industrial Anomaly Detection Multimodal Anomaly Detection
Corporate Research, Robert Bosch GmbH, Renningen, Germany · Otto-von-Guericke-University Magdeburg, Magdeburg, Germany · Institute for Artificial Intelligence, University of Stuttgart, Stuttgart, Germany +2
AD-FM: Multimodal LLMs for Anomaly Detection via Multi-Stage Reasoning and Fine-Grained Reward Optimization
Aug 6, 2025Jingyi Liao, Yongyi Su, Rong-Cheng Tu +6VLM Adaptation Industrial Anomaly Detection
1Nanyang Technological University · Institute for Infocomm Research (I2R), A*STAR · 3South China University of Technology
Uncertainty-Weighted Fusion of Image and Synthetic Event for Video Anomaly Detection
May 5, 2025Sungheon Jeong, Jihong Park, Mohsen ImaniUncertainty-Aware Multimodal Fusion Multimodal Anomaly Detection
University of California, Irvine · MOLOCO
Tuned Reverse Distillation: Enhancing Multimodal Industrial Anomaly Detection with Crossmodal Tuners
Dec 12, 2024Xinyue Liu, Jianyuan Wang, Biao Leng +1Cross-Modal Knowledge Distillation Industrial Anomaly Detection
School of Computer Science and Engineering, Beihang University, Beijing 100191, China · Key Laboratory of Intelligent Bionic Unmanned Systems, Ministry of Education, and the School of Artificial Intelligence, University of Science and Technology Beijing, Beijing 100083, China · School of Computer Science and Technology, Beijing Jiaotong University, Beijing 100044, China

Feb 19, 2026J/K move · Enter open · S save

EAGLE: Expert-Augmented Attention Guidance for Tuning-Free Industrial Anomaly Detection in Multimodal Large Language Models

Xiaomeng Peng, Xilang Huang, Seon Han Choi

Ewha Womans University

Anomaly Detection Industrial Anomaly Detection Multimodal Anomaly Detection Multimodal Large Language Models

Open paper AskPDF arXiv

Multimodal large language models (MLLMs) can enrich industrial anomaly detection with semantic descriptions and anomaly reasoning, but they still lag specialist anomaly detectors in binary detection accuracy. Existing approaches address this gap by fine-tuning MLLMs or training bridging modules to align expert outputs with MLLM inputs, limiting flexibility across backbones. We propose EAGLE, a tuning-free framework that integrates expert anomaly detectors with frozen MLLMs. EAGLE consists of Threshold-Guided Prompt Selection (TGPS), which estimates a decision threshold from expert model statistics and selects textual and visual prompts, and Confidence-Aware Attention Sharpening (CAAS), which shifts MLLM attention toward visual evidence when expert confidence is low. Beyond improving accuracy, we analyze MLLM attention and find that correct anomaly predictions are associated with stronger focus on ground-truth defect regions; EAGLE consistently strengthens this alignment. On MVTec-AD and VisA, EAGLE improves five MLLM backbones without parameter updates, reaching up to 94.4% and 88.1% in anomaly discrimination accuracy, respectively, and achieving performance competitive with fine-tuning-based methods while largely preserving MLLM semantic reasoning ability.