Application of a Self-supervised Learning Technique for Monitoring Industrial Spaces

Supervised learning has reached a bottleneck as they require expensive and time-consuming annotations. In addition, in some problems, such as in industrial spaces, it is not always possible to acquire a large number of images. Self-supervised learning helps these issues by extracting information from the data itself, without requiring labels and has achieved good performance, closing the gap between supervised and self-supervised learning. This work presents the application of a self-supervised learning method - SwAV, that classifies anomalies in an industrial space, evaluates its performance and compares the results to the supervised paradigm.