Please use this identifier to cite or link to this item:
Title: Human activity recognition based on hidden Markov models
Authors: Liu, Xiao Hui
Keywords: DRNTU::Engineering::Electrical and electronic engineering::Electronic systems::Signal processing
DRNTU::Engineering::Electrical and electronic engineering::Control and instrumentation::Control engineering
Issue Date: 2006
Source: Liu, X. H. (2006). Human activity recognition based on hidden Markov models. Doctoral thesis, Nanyang Technological University, Singapore.
Abstract: This thesis discusses the main issues of human activity recognition systems, including automatic human activity segmentation, non-meaningful activity rejection and multi-agent activity recognition, and presents the contribution of this project for these issues. Three contributions are presented in this thesis. Firstly, a background-state based auto-segmentation framework is proposed to segment human activities of interest from continuous input. Secondly, the non-meaningful activities is rejected be a pairwise likelihood ratio test (PLRT), which has a good performance while only relying on information of meaningful patterns. Thirdly, an observation decomposed hidden Markov model (ODHMM) is proposed to recognize multi-agent activities, where the role of each agent can be identified automatically. These contributions concerned on various important aspects of human activity recognition and make it possible to build a real-life system.
DOI: 10.32657/10356/4747
Rights: Nanyang Technological University
Fulltext Permission: open
Fulltext Availability: With Fulltext
Appears in Collections:EEE Theses

Files in This Item:
File Description SizeFormat 
EEE-THESES_750.pdf7.87 MBAdobe PDFThumbnail

Page view(s) 50

Updated on May 9, 2021

Download(s) 5

Updated on May 9, 2021

Google ScholarTM




Items in DR-NTU are protected by copyright, with all rights reserved, unless otherwise indicated.