Deep Learning

Supervised Domain Adaptation for Surface Defect Detection Leveraging Partial Data Availability

Novel domain adaptation approach for surface defect detection that handles partial data availability using modified triplet loss and two-stage training.