Model-based Transfer Learning for Automatic Optical Inspection based on domain discrepancy

TL;DR

This paper introduces a model-based transfer learning approach for automatic optical inspection that leverages domain similarity and data augmentation to enhance performance, achieving up to 20% improvements in F1 score.

Contribution

It proposes a novel domain discrepancy-based transfer learning method that considers domain similarity and data imbalance for improved AOI performance.

Findings

F1 score and PR curve improved by up to 20%

Domain discrepancy score effectively identifies related source datasets

Data augmentation reduces class imbalance in target and source domains

Abstract

Transfer learning is a promising method for AOI applications since it can significantly shorten sample collection time and improve efficiency in today's smart manufacturing. However, related research enhanced the network models by applying TL without considering the domain similarity among datasets, the data long-tailedness of a source dataset, and mainly used linear transformations to mitigate the lack of samples. This research applies model-based TL via domain similarity to improve the overall performance and data augmentation in both target and source domains to enrich the data quality and reduce the imbalance. Given a group of source datasets from similar industrial processes, we define which group is the most related to the target through the domain discrepancy score and the number of samples each has. Then, we transfer the chosen pre-trained backbone weights to train and fine-tune the target network. Our research suggests increases in the F1 score and the PR curve up to 20% compared with TL using benchmark datasets.