Few-shot multi-view object classification via dual augmentation network

Yaqian Zhou, Haochun Lu, Tong Hao, Xuanya Li, An-An Liu

Abstract: Highlights•We explore a few-shot multi-view object classification (FS-MVOC) task.•A dual augmentation network is developed for the FS-MVOC task.•A multi-view representation extractor is pre-trained in a meta-learning way.•Attention-guided multi-view representation augmentation is employed.•Category prototype augmentation is adopted to obtain representative prototypes.