Deep multi-path convolutional neural network joint with salient region attention for facial expression recognition

作者：

Highlights：

• A deep-based model for solving facial expression recognition is proposed.

• Propose a Salient Expression Region Descriptor (SERD) to locate expression-sensitive regions and refine deep features.

• Construct an encoder-decoder framework to disentangle multiple variations (e.g. ages, races, poses, etc.) from features.

• Experiments on six famous expression databases verify the effectiveness of the proposed model.

摘要

•A deep-based model for solving facial expression recognition is proposed.•Propose a Salient Expression Region Descriptor (SERD) to locate expression-sensitive regions and refine deep features.•Construct an encoder-decoder framework to disentangle multiple variations (e.g. ages, races, poses, etc.) from features.•Experiments on six famous expression databases verify the effectiveness of the proposed model.

论文关键词：Attention,Convolutional neural network,Facial expression recognition,Multi-Path variation-suppressing network,Salient expressional region descriptor

论文评审过程：Received 22 June 2018, Revised 19 February 2019, Accepted 21 March 2019, Available online 28 March 2019, Version of Record 4 April 2019.

论文官网地址：https://doi.org/10.1016/j.patcog.2019.03.019