Novel Class Discovery

Novel Class Discovery (NCD) is a crucial task in the field of computer vision, aiming to identify new categories within unlabelled data by leveraging prior knowledge of known categories. This task divides the dataset into two parts: a labelled dataset of known categories and an unlabelled dataset of unknown categories. The core objective of NCD is to enhance the model's ability to discover new categories, thereby improving the system's generalization performance and adaptability, which holds significant application value.