A Time Flexible Kernel framework for video-based activity recognition

作者：

Highlights：

• TFK: a kernel framework between arbitrary length sequences.

• Some complex activities are defined by the order of sub-actions.

• The new kernel framework improves results in complex activities recognition.

• Combination of several levels of granularity in temporal divisions reduces clutter.

摘要

•TFK: a kernel framework between arbitrary length sequences.•Some complex activities are defined by the order of sub-actions.•The new kernel framework improves results in complex activities recognition.•Combination of several levels of granularity in temporal divisions reduces clutter.

论文关键词：Activity recognition,Soft-assignment,Kernel methods,Support Vector Machine

论文评审过程：Received 21 January 2015, Revised 4 September 2015, Accepted 23 December 2015, Available online 24 February 2016, Version of Record 25 March 2016.

论文官网地址：https://doi.org/10.1016/j.imavis.2015.12.006