Deep Learning and XAI Techniques for Anomaly Detection

5

Computer Vision Anomaly Explainability

Computer vision anomaly detection has broad applications, such as industrial manufacturing, medical imaging analysis, and autonomous driving. Detecting anomalies in image data is challenging yet crucial to identifying abnormal events, safety risks, and quality defects that may cause financial and brand damage to many companies.

According to the American Society for Quality at https://asq.org/quality-resources/cost-of-quality, organizations can incur 15-20% quality-related costs from their sales revenue. Besides financial loss, quality-related activities can result in poor customer satisfaction, potential lawsuits, and productivity impact. Hence, research for visual anomaly detection plays a critical role ...