Self-reconstruction network for fine-grained few-shot classification

Xiaoxu Li, Zhen Li, Jiyang Xie, Xiaochen Yang, Jing-Hao Xue, Zhanyu Ma

Abstract: Highlights•A novel fine-grained few-shot classification method is proposed to alleviate overfitting.•The proposed method uses self-reconstruction to increase the diversity of query features.•The proposed method introduces a restrained cross-entropy loss.•The proposed method generalizes well across five benchmark datasets.