PANDAS: Prototype-based Novel Class Discovery and Detection

arXiv:2402.17420v1 Announce Type: new
Abstract: Object detectors are typically trained once and for all on a fixed set of classes. However, this closed-world assumption is unrealistic in practice, as new classes will inevitably emerge after the detector is deployed in the wild. In this work, we look at ways to extend a detector trained for a set of base classes so it can i) spot the presence of novel classes, and ii) automatically enrich its repertoire to be able to …

abstract arxiv class cs.ai cs.cv detection discovery look novel pandas practice set type will work world