JUDO: A Juxtaposed Domain-Oriented Multimodal Reasoner for Industrial Anomaly QA

2026-05-19

AI Summary

This work addresses the challenge that existing large-scale multimodal models struggle to accurately answer anomaly-related questions in complex industrial scenarios due to a lack of domain-specific knowledge. To overcome this limitation, the authors propose a juxtaposed domain-context mechanism that constructs fine-grained visual contrasts by placing query images alongside normal reference images. Furthermore, they introduce a domain-oriented reasoning pipeline that integrates supervised fine-tuning with GRPO-based reinforcement learning to effectively inject industrial prior knowledge into the model. Evaluated on the MMAD benchmark, the proposed approach significantly outperforms Qwen2.5-VL-7B and GPT-4o, demonstrating that combining visual contrast with domain-specific knowledge substantially enhances the accuracy of industrial anomaly question answering.

Abstract

Industrial anomaly detection has been significantly advanced by Large Multimodal Models (LMMs), enabling diverse human instructions beyond detection, particularly through visually grounded reasoning for better image understanding. However, LMMs lack domain-specific knowledge, which limits their ability to generate accurate responses in complex industrial scenarios. In this work, we present JUDO, Juxtaposed Domain-Oriented Multimodal Reasoner, a framework that efficiently incorporates domain knowledge and context in visual and textual reasoning. Through visual reasoning, our model segments the defect region by juxtaposing query images with normal images as visual domain context, enabling a fine-grained visual comparative inspection. Furthermore, we inject domain knowledge through supervised fine-tuning (SFT) to enhance context understanding and subsequently guide domain reasoning through reinforcement learning (GRPO) with tailored rewards, opting for a domain-oriented reasoning process. Experimental results demonstrate that JUDO achieves superior performance on the MMAD benchmark, surpassing models such as Qwen2.5-VL-7B and GPT-4o. These results highlight the importance of enhancing domain knowledge and context for effective reasoning in anomaly understanding.

Problem

• Industrial anomaly detection
• Domain-specific knowledge
• Multimodal reasoning
• Visual grounding
• Context understanding

Innovation

• domain-oriented reasoning
• visual juxtaposition
• multimodal anomaly detection
• reinforcement learning with tailored rewards
• supervised fine-tuning