Semi-supervised anomaly detection with dual prototypes autoencoder for industrial surface inspection

• A semi-supervised anomaly detection method, dual prototype autoencoder (DPAE), is proposed to distinguish anomalies on the industrial products surface. • The dual prototype loss in DPAE can reduce the intra-class distances of normal samples, resulting in a more separable boundary between the defective and defect-free images. • The aluminum profile surface defect (APSD) dataset is constructed in this work, which includes the defect-free samples and the defective images from ten defect types. • The proposed DPAE can achieve promising results when the defective samples are not available. • This work provides a new way to apply convolutional auto-encoder in data-limited industrial inspection tasks. Anomaly detection in the automated optical quality inspection is of great important for guaranteeing the surface quality of industrial products. Most related methods are based on supervised learning techniques, which require a large number of normal and anomalous samples to obtain a robust classifier. However, the diversity of potential defects and low availability of defective samples during manufacturing bring more challenges to anomaly detection. Based on the encoder-decoder-encoder paradigm, a semi-supervised anomaly detection method Dual Prototype Auto-Encoder (DPAE) is proposed in this paper. At the training stage, the dual prototype loss and reconstruction loss are introduced to encourage the latent vectors generated by the encoders to keep closer to their own prototype. Therefore, two latent vectors of the normal image tend to be closer, and large distance between the latent vectors indicates an anomaly. And we also construct the Aluminum Profile Surface Defect (APSD) dataset for the anomaly detection task. Finally, extensive experiments on four datasets show that DPAE is effective and outperforms state-of-the-art methods.