Industrial Inspection
Overview
Industrial inspection studies how to find, locate, measure, and assess defects in manufactured products and infrastructure, using images and sensors such as RGB-D, thermal, X-ray, ultrasound, and radar. Common approaches include object detection and segmentation with CNNs and transformers, anomaly detection trained on normal examples, and multimodal or vision-language models that connect visual evidence to descriptions or diagnoses. Self-supervised learning, synthetic defect generation, and foundation-model adaptation help when labeled defect examples are scarce, while robots and edge systems bring inspection into factories and field settings. A central challenge is making results reliable beyond curated benchmarks: real deployments face changing materials, viewpoints, and imaging conditions, so work increasingly tests generalization, measurement accuracy, human review, and safety alongside detection scores.
Latest papers
120 papers12 papers in the last four weeks, down 14% on the four weeks before. 0.1% of all new papers.