Placing Objects in Context via Inpainting for Out-of-distribution Segmentation

arXiv:2402.16392v1 Announce Type: new
Abstract: When deploying a semantic segmentation model into the real world, it will inevitably be confronted with semantic classes unseen during training. Thus, to safely deploy such systems, it is crucial to accurately evaluate and improve their anomaly segmentation capabilities. However, acquiring and labelling semantic segmentation data is expensive and unanticipated conditions are long-tail and potentially hazardous. Indeed, existing anomaly segmentation datasets capture a limited number of anomalies, lack realism or have strong domain shifts. In …

abstract anomaly arxiv capabilities context cs.cv data deploy distribution inpainting labelling objects segmentation semantic systems training type via will world