Three Stream 3D CNN with SE Block for Micro-Expression Recognition

Micro-expression is a kind of short-term and imperceptible facial expression. In this paper, we propose a method of micro-expression recognition using SEnet and three-stream 3D CNN (SETFNet). Three different facial regions are used as the input of three streams. Adding SE block to the network can learn the weights of each feature channel adaptively. Experimental results show that this method can effectively improve the recognition performance of micro-expression.