Changes

This Technical Note (TN for short) belongs to the series introduced [[ML-TN-003 — AI at the edge: visual inspection of assembled PCBs for defect detection — Part 1|herethis series of articles]]. Specifically, it illustrates the attempt of employing ML and image processing techniques in the field of automatic visual inspection of Printed Circuit Boards (PCB-AVI) for building a proprietary dataset of SMD-populated PCBs exhibiting mounting anomalies. This real-world use case is affected by the typical problem ML engineers and data scientists need to face when it comes to the detection of defects/anomalies related to an industrial process: these defects/anomalies are extremely rare! As a result, no large datasets of samples are available to train the models.