A novel micro-expression detection algorithm based on BERT and 3DCNN

作者：

Highlights：

• We utilize candidate segment generation and spatio-temporal feature extraction modules to obtain spatio-temporal features.

• The grouping module merges consecutive segments and suppresses overlapping segments.

• Comprehensive experiments confirmed the effectiveness of the proposed method.

摘要

•We utilize candidate segment generation and spatio-temporal feature extraction modules to obtain spatio-temporal features.•The grouping module merges consecutive segments and suppresses overlapping segments.•Comprehensive experiments confirmed the effectiveness of the proposed method.

论文关键词：Micro-expression detection,3DCNN,BERT

论文评审过程：Received 15 September 2021, Revised 9 December 2021, Accepted 13 January 2022, Available online 31 January 2022, Version of Record 7 February 2022.

论文官网地址：https://doi.org/10.1016/j.imavis.2022.104378