Human activity recognition using temporal convolutional neural network architecture

作者：

Highlights：

• Temporal convolutional neural network for leveraging spatio-temporal features.

• Reduce dataset size using only short videos from one typical camera.

• Proposed architecture recognizes human activity and shares the result to NAO robot.

• Results demonstrate that the proposed classifier can be used to HAR in real time.

摘要

•Temporal convolutional neural network for leveraging spatio-temporal features.•Reduce dataset size using only short videos from one typical camera.•Proposed architecture recognizes human activity and shares the result to NAO robot.•Results demonstrate that the proposed classifier can be used to HAR in real time.

论文关键词：Human activity recognition,3D convolution,Video processing,Temporal CNN

论文评审过程：Received 26 March 2021, Revised 24 August 2021, Accepted 22 November 2021, Available online 8 December 2021, Version of Record 11 December 2021.

论文官网地址：https://doi.org/10.1016/j.eswa.2021.116287