A Multiscale Approach to Detect the Semantics of Locomotion without Positioning Information

Chen Qu, Jinxin Yang, Guangwen Si, Yufei Zhao, Yiren Zhou, Wen Chi Yang*

*Corresponding author for this work

Research output: Chapter in Book or Report/Conference proceedingConference Proceedingpeer-review

Abstract

The detection of locomotion patterns and trajectory semantics has long relied on the positioning information. However, positioning information can be unavailable in specific scenarios. In addition, the position values are sensitive to rotation and hence a bias can exist during the training process in deep learning models. This paper introduces an alternative model that classifies trajectories based on accelerometers, instead of positioning systems. We built up a convolutional neural network that inputs the degree of velocity and turning angles at several time scales and converts this information into a semantic class. We examined the model in a simulated environment and also a benchmark task. This model has exhibited a competitive performance even compared with those models based on positioning information.

Original languageEnglish
Title of host publication2020 7th International Conference on Soft Computing and Machine Intelligence, ISCMI 2020
PublisherInstitute of Electrical and Electronics Engineers Inc.
Pages270-274
Number of pages5
ISBN (Electronic)9781728175591
DOIs
Publication statusPublished - 14 Nov 2020
Externally publishedYes
Event7th International Conference on Soft Computing and Machine Intelligence, ISCMI 2020 - Virtual, Stockholm, Sweden
Duration: 14 Nov 202015 Nov 2020

Publication series

Name2020 7th International Conference on Soft Computing and Machine Intelligence, ISCMI 2020

Conference

Conference7th International Conference on Soft Computing and Machine Intelligence, ISCMI 2020
Country/TerritorySweden
CityVirtual, Stockholm
Period14/11/2015/11/20

Keywords

  • acceleration sensors
  • deep learning
  • locomotion semantics
  • rotational invariance
  • trajectory classification

Fingerprint

Dive into the research topics of 'A Multiscale Approach to Detect the Semantics of Locomotion without Positioning Information'. Together they form a unique fingerprint.

Cite this