Evaluating Modern Visual Anomaly Detection Approaches in Semiconductor Manufacturing: A Comparative Study

TL;DR

This paper benchmarks modern unsupervised visual anomaly detection methods in semiconductor manufacturing, demonstrating their effectiveness in analyzing SEM images and aiding defect identification without requiring labeled anomalies.

Contribution

It introduces a new benchmark for VAD in semiconductor manufacturing using the MIIC dataset, enabling systematic evaluation of current approaches.

Findings

Modern VAD methods are effective in semiconductor SEM image analysis.

The benchmark facilitates comparison of different VAD techniques.

Results show promising potential for unsupervised defect detection.

Abstract

Semiconductor manufacturing is a complex, multistage process. Automated visual inspection of Scanning Electron Microscope (SEM) images is indispensable for minimizing equipment downtime and containing costs. Most previous research considers supervised approaches, assuming a sufficient number of anomalously labeled samples. On the contrary, Visual Anomaly Detection (VAD), an emerging research domain, focuses on unsupervised learning, avoiding the costly defect collection phase while providing explanations of the predictions. We introduce a benchmark for VAD in the semiconductor domain by leveraging the MIIC dataset. Our results demonstrate the efficacy of modern VAD approaches in this field.