Multi-Modal Anomaly Navigation (ANNA)

Abstract

Navigation in natural outdoor environments requires a robust and reliable traversability classification method to handle the plethora of situations a robot can encounter. Binary classification algorithms perform well in their native domain but tend to provide overconfident predictions when presented with out-of-distribution samples, which can lead to catastrophic failure when navigating unknown environments. We propose to overcome this issue by using anomaly detection on multi-modal images for traversability classification, which is easily scalable by training in a self-supervised fashion from robot experience. This dataset serves to validate anomaly detection algorithms for their performance on multi-modal image data. Positive training samples are generated by projecting footholds of a walking robot into camera images and extracting image patches around them. Both positive and negative validation data was labelled by a human expert.