Multi-stage adaptive regression for online activity recognition

作者：

Highlights：

• An effective multi-stage adaptive regression framework is proposed to address the partial activity observation problem in online activity recognition.

• Multiple score functions are collaboratively learned via an adaptive label strategy to reinforce the capacity of distinguishing similar partial activities and the robustness to arbitrary activity fragments.

• A challenging interaction database, Online Human Interaction (OHI), is collected in a realistic scenario to further evaluate the online activity recognition.

摘要

•An effective multi-stage adaptive regression framework is proposed to address the partial activity observation problem in online activity recognition.•Multiple score functions are collaboratively learned via an adaptive label strategy to reinforce the capacity of distinguishing similar partial activities and the robustness to arbitrary activity fragments.•A challenging interaction database, Online Human Interaction (OHI), is collected in a realistic scenario to further evaluate the online activity recognition.

论文关键词：Online activity recognition,Interaction recognition,Partial observation,Adaptive regression

论文评审过程：Received 8 October 2018, Revised 16 June 2019, Accepted 12 September 2019, Available online 13 September 2019, Version of Record 3 October 2019.

论文官网地址：https://doi.org/10.1016/j.patcog.2019.107053