Abstract: Deploying deep learning models for industrial defect detection is challenging, as smaller models lack precision and larger models require significant computational resources. To address this ...