Exploiting inter-frame regional correlation for efficient action recognition

作者：

Highlights：

• We proposed a novel temporal feature extraction method for action recognition.

• The novel method explores inter-frame correlation on the regional level.

• Our method achieves state-of-the-art performance on benchmark datasets.

摘要

•We proposed a novel temporal feature extraction method for action recognition.•The novel method explores inter-frame correlation on the regional level.•Our method achieves state-of-the-art performance on benchmark datasets.

论文关键词：Action recognition,Inter-frame correlation,Feature extraction

论文评审过程：Received 2 December 2019, Revised 16 December 2020, Accepted 1 March 2021, Available online 11 March 2021, Version of Record 19 April 2021.

论文官网地址：https://doi.org/10.1016/j.eswa.2021.114829