Discriminative Feature Enhancement Network for few-shot classification and beyond

Fangyu Wu, Qiufeng Wang, Xuan Liu, Qi Chen, Yuxuan Zhao, Bailing Zhang, Eng Gee Lim

Abstract: Highlights•DFENet introduces a novel approach for few-shot classification, using visual and semantic knowledge.•The neural decoding-based attention module efficiently assigns attention weights to keys, inspired by cognitive neuroscience.•A flexible triplet loss optimizes class pair margins based on semantic similarity.•DFENet achieves state-of-the-art results in few-shot classification, facial expression recognition, and image retrieval tasks.