LocalFace: Learning significant local features for deep face recognition

作者：

Highlights：

• Visualization and statistics of significant features in face images.

• A novel face local feature extraction method to ensemble local features into global features.

• A novel face random occlusion data augmentation which can better simulate the occlusion in real scenes.

摘要

•Visualization and statistics of significant features in face images.•A novel face local feature extraction method to ensemble local features into global features.•A novel face random occlusion data augmentation which can better simulate the occlusion in real scenes.

论文关键词：Face recognition,ArcFace,Channel attention,Deep learning

论文评审过程：Received 20 May 2021, Revised 1 January 2022, Accepted 9 May 2022, Available online 16 May 2022, Version of Record 21 May 2022.

论文官网地址：https://doi.org/10.1016/j.imavis.2022.104484