This thesis addresses deep learning-based methods for automatic anomaly detection in an industrial context. It involves image- or sensor-based detection of defects in the production process that can affect the quality of ...