Composite Score for Anomaly Detection in Imbalanced Real-World Industrial Dataset
This addresses quality control in manufacturing by improving anomaly detection under strict constraints, though it is incremental as it builds on existing methods.
The paper tackled anomaly detection in imbalanced industrial datasets, such as Printed Circuit Board Assembly images, by using a VQGAN-based reconstruction and multi-level metrics to achieve a composite score, resulting in 95.69% accuracy on MVTec-AD and 87.93% under zero-false-negative constraints on a partner dataset.
In recent years, the industrial sector has evolved towards its fourth revolution. The quality control domain is particularly interested in advanced machine learning for computer vision anomaly detection. Nevertheless, several challenges have to be faced, including imbalanced datasets, the image complexity, and the zero-false-negative (ZFN) constraint to guarantee the high-quality requirement. This paper illustrates a use case for an industrial partner, where Printed Circuit Board Assembly (PCBA) images are first reconstructed with a Vector Quantized Generative Adversarial Network (VQGAN) trained on normal products. Then, several multi-level metrics are extracted on a few normal and abnormal images, highlighting anomalies through reconstruction differences. Finally, a classifer is trained to build a composite anomaly score thanks to the metrics extracted. This three-step approach is performed on the public MVTec-AD datasets and on the partner PCBA dataset, where it achieves a regular accuracy of 95.69% and 87.93% under the ZFN constraint.