Hierarchical Complex Activity Representation and Recognition Using Topic Model and Classifier Level Fusion
Authors:
- 31 August 2016
- journal article
- research article
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Biomedical Engineering
- Vol. 64 (6) , 1369-1379
Abstract
Human activity recognition is an important area of ubiquitous computing. Most current researches in activity recognition mainly focus on simple activities, e.g., sitting, running, walking, and standing. Compared with simple activities, complex activities are more complicated with high-level semantics, e.g., working, commuting, and having a meal. This paper presents a hierarchical model to recognize complex activities as mixtures of simple activities and multiple actions. We generate the components of complex activities using a clustering algorithm, represent and recognize complex activities by applying a topic model on these components. It is a data-driven method that can retain effective information for representing and recognizing complex activities. In addition, acceleration and physiological signals are fused in classifier level to ensure the overall performance of complex activity recognition. The results of experiments show that our method has ability to represent and recognize complex activities effectively.Keywords
Funding Information
- Ministry of Industry and Information Technology of China (2010ZX01042-002-003-001)
- China Knowledge Centre for Engineering Sciences and Technology (CKCEST-2014-1-5)
- National Natural Science Foundation of China (60703040, 61332017)
- Science and Technology Department of Zhejiang Province (2011C13042, 2015C33002)
This publication has 27 references indexed in Scilit:
- Smartphone‐based activity recognition independent of device orientation and placementInternational Journal of Communication Systems, 2015
- Behavior analysis of elderly using topic modelsPervasive and Mobile Computing, 2014
- Discovery of activity composites using topic models: An analysis of unsupervised methodsPervasive and Mobile Computing, 2014
- Feature Selection and Activity Recognition System Using a Single Triaxial AccelerometerIEEE Transactions on Biomedical Engineering, 2014
- Attribute-Based Classification for Zero-Shot Visual Object CategorizationIEEE Transactions on Pattern Analysis and Machine Intelligence, 2013
- Centinela: A human activity recognition system based on acceleration and vital sign dataPervasive and Mobile Computing, 2012
- The WEKA data mining softwareACM SIGKDD Explorations Newsletter, 2009
- A novel sequence representation for unsupervised analysis of human activitiesArtificial Intelligence, 2009
- Daily Routine Recognition through Activity Spotting2009
- Unsupervised Learning of Human Action Categories Using Spatial-Temporal WordsInternational Journal of Computer Vision, 2008