SIAD: Self-supervised Image Anomaly Detection System
This addresses the need for long-term, automated visual inspection in manufacturing automation, though it appears incremental as it combines existing self-supervised and supervised methods.
The paper tackles the problem of automating visual inspection in manufacturing by introducing SIAD, a self-supervised system that uses anomaly-free data to generate labels and train supervised models for continuous online application, with deployment in real-life industrial settings.
Recent trends in AIGC effectively boosted the application of visual inspection. However, most of the available systems work in a human-in-the-loop manner and can not provide long-term support to the online application. To make a step forward, this paper outlines an automatic annotation system called SsaA, working in a self-supervised learning manner, for continuously making the online visual inspection in the manufacturing automation scenarios. Benefit from the self-supervised learning, SsaA is effective to establish a visual inspection application for the whole life-cycle of manufacturing. In the early stage, with only the anomaly-free data, the unsupervised algorithms are adopted to process the pretext task and generate coarse labels for the following data. Then supervised algorithms are trained for the downstream task. With user-friendly web-based interfaces, SsaA is very convenient to integrate and deploy both of the unsupervised and supervised algorithms. So far, the SsaA system has been adopted for some real-life industrial applications.