Human activity recognition using multi-features and multiple kernel learning

作者：

Highlights：

• Presenting 3D shape features based on spherical harmonics representation.

• Presenting 3D motion features of the distal limb segments using kinematic structure.

• Fusing multiple features using multi kernel learning for human action recognition.

摘要

Highlights•Presenting 3D shape features based on spherical harmonics representation.•Presenting 3D motion features of the distal limb segments using kinematic structure.•Fusing multiple features using multi kernel learning for human action recognition.

论文关键词：Human activity recognition,Multiple kernel learning,Spherical harmonics coefficients,Distal limb segments,Support vector machines

论文评审过程：Received 17 April 2013, Revised 8 November 2013, Accepted 30 November 2013, Available online 17 December 2013.

论文官网地址：https://doi.org/10.1016/j.patcog.2013.11.032