Occlusion expression recognition based on non-convex low-rank double dictionaries and occlusion error model

作者：

Highlights：

• Using non-convex logarithm low-rank decomposition to separate expression feature from identity feature to decrease the expression intra-class differences.

• The expression dictionary is used alone to sparsely represent the expression feature of the original image for recognition.

• Occluded error model is defined to represent the error matrix of occlusion information.

摘要

•Using non-convex logarithm low-rank decomposition to separate expression feature from identity feature to decrease the expression intra-class differences.•The expression dictionary is used alone to sparsely represent the expression feature of the original image for recognition.•Occluded error model is defined to represent the error matrix of occlusion information.

论文关键词：Non-convex,Low-rank decomposition,Expression feature,Individual identity features,Occlusion error model

论文评审过程：Received 12 November 2018, Revised 11 March 2019, Accepted 9 April 2019, Available online 22 April 2019, Version of Record 4 May 2019.

论文官网地址：https://doi.org/10.1016/j.image.2019.04.006