Fall detection without people: A simulation approach tackling video data scarcity

作者：

Highlights：

• Fall detection system based on a myoskeletal (physics-based) simulation.

• No need for video recordings of human falls.

• Persons height is used to parameterise the simulation, addressing human variability.

• State-of-the-art performance, tested in publicly available datasets.

• System is robust for up to 50% occlusion.

摘要

•Fall detection system based on a myoskeletal (physics-based) simulation.•No need for video recordings of human falls.•Persons height is used to parameterise the simulation, addressing human variability.•State-of-the-art performance, tested in publicly available datasets.•System is robust for up to 50% occlusion.

论文关键词：Fall detection,Physics simulation,Depth video,Visual occlusions,Myoskeletal modelling,Assisted living

论文评审过程：Received 7 December 2017, Revised 25 April 2018, Accepted 7 June 2018, Available online 13 June 2018, Version of Record 26 June 2018.

论文官网地址：https://doi.org/10.1016/j.eswa.2018.06.019