Paper Title: Enhanced Anomaly Detection in Advanced Manufacturing Execution using YOLO V8
Conference Name: KICS 2024
Abstract: The experiment in this study highlight two cate-
gories of anomalies which we call structural and logical anoma-
lies. Anomaly detection is crucial to ensuring the robustness and
reliability of operational systems, preventing potential failures
and enhancing decision-making processes. The effectiveness of the
YOLO v8 anomaly detection model in identifying these classes of
anomalies in manufacturing systems shows the model’s superior
performance in detecting and classifying defects. Dataset used
here is made up of 3644 images divided into good, logical and
structural anomalies. To the best of our knowledge, previous
works predominantly concentrated on the development of meth-
ods for the detection of dents and scratches. We therefore created
a new method for the supervised localization of anomalies that
focuses on the detection of both structural and logical anomalies.