Zero-Shot Anomaly Detection in Battery Thermal Images Using Visual Question Answering with Prior Knowledge

2025-05-22

AI Summary

Anomaly detection in battery thermal imaging faces challenges of scarce anomalous samples, high annotation costs, and safety risks associated with labeling. Method: This paper proposes a zero-shot visual question answering (VQA) framework that requires no annotated battery anomaly data. Leveraging large multimodal models—including ChatGPT-4o, LLaVA-13b, and BLIP-2—the approach employs prior-knowledge-driven textual prompting to encode normal thermal behavior as interpretable semantic constraints, enabling generalizable anomaly identification without model fine-tuning. Contribution/Results: To our knowledge, this is the first work to introduce zero-shot VQA into battery thermal safety diagnostics. Comprehensive multi-round robustness evaluations confirm reliability. Experiments demonstrate that the method achieves detection performance on unseen battery data comparable to state-of-the-art supervised methods, significantly enhancing both practicality and interpretability in industrial thermal safety assessment.

Technology Category

• Computer Vision: Multi-modal Vision
• Machine Learning: Large Multimodal Models (LMMs)
• Natural Language Processing: Safety and Robustness

Application Category

• Search and Retrieval-Augmented AI: Web question answering
• Graph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs
• Security and Privacy: Large-scale security measurements

Abstract

Batteries are essential for various applications, including electric vehicles and renewable energy storage, making safety and efficiency critical concerns. Anomaly detection in battery thermal images helps identify failures early, but traditional deep learning methods require extensive labeled data, which is difficult to obtain, especially for anomalies due to safety risks and high data collection costs. To overcome this, we explore zero-shot anomaly detection using Visual Question Answering (VQA) models, which leverage pretrained knowledge and textbased prompts to generalize across vision tasks. By incorporating prior knowledge of normal battery thermal behavior, we design prompts to detect anomalies without battery-specific training data. We evaluate three VQA models (ChatGPT-4o, LLaVa-13b, and BLIP-2) analyzing their robustness to prompt variations, repeated trials, and qualitative outputs. Despite the lack of finetuning on battery data, our approach demonstrates competitive performance compared to state-of-the-art models that are trained with the battery data. Our findings highlight the potential of VQA-based zero-shot learning for battery anomaly detection and suggest future directions for improving its effectiveness.

Problem

• Detect battery thermal anomalies without labeled training data
• Leverage VQA models for zero-shot anomaly detection
• Evaluate robustness of VQA models in battery safety applications

Innovation

• Zero-shot anomaly detection using VQA models
• Leveraging pretrained knowledge with text prompts
• Detecting anomalies without battery-specific training data