Treshold dynamic time warping for spatial activity recognition

Non-invasive spatial activity recognition is a difficult task, complicated by variation in how the same activities are conducted and furthermore by noise introduced by video tracking procedures. In this paper we propose an algorithm based on dynamic time warping (DTW) as a viable method with which t...

Full description

Bibliographic Details
Main Authors: Riedel, Daniel, Venkatesh, Svetha, Liu, Wan-Quan
Format: Journal Article
Published: Institute of Scientific Computing and Information 2007
Online Access:http://www.math.ualberta.ca/ijiss/SS-volume-3-07.htm
http://hdl.handle.net/20.500.11937/7381
_version_ 1848745352277000192
author Riedel, Daniel
Venkatesh, Svetha
Liu, Wan-Quan
author_facet Riedel, Daniel
Venkatesh, Svetha
Liu, Wan-Quan
author_sort Riedel, Daniel
building Curtin Institutional Repository
collection Online Access
description Non-invasive spatial activity recognition is a difficult task, complicated by variation in how the same activities are conducted and furthermore by noise introduced by video tracking procedures. In this paper we propose an algorithm based on dynamic time warping (DTW) as a viable method with which to quantify segmented spatial activity sequences from a video tracking system. DTW is a widely used technique for optimally aligning or warping temporal sequences through minimisation of the distance between their components. The proposed algorithm threshold DTW (TDTW) is capable of accurate spatial sequence distance quantification and is shown using a three class spatial data set to be more robust and accurate than DTW and the discrete hidden markov model (HMM). We also evaluate the application of a band dynamic programming (DP) constraint to TDTW in order to reduce extraneous warping between sequences and to reduce the computation complexity of the approach. Results show that application of a band DP constraint to TDTW improves runtime performance significantly, whilst still maintaining a high precision and recall.
first_indexed 2025-11-14T06:15:59Z
format Journal Article
id curtin-20.500.11937-7381
institution Curtin University Malaysia
institution_category Local University
last_indexed 2025-11-14T06:15:59Z
publishDate 2007
publisher Institute of Scientific Computing and Information
recordtype eprints
repository_type Digital Repository
spelling curtin-20.500.11937-73812017-01-30T10:59:30Z Treshold dynamic time warping for spatial activity recognition Riedel, Daniel Venkatesh, Svetha Liu, Wan-Quan Non-invasive spatial activity recognition is a difficult task, complicated by variation in how the same activities are conducted and furthermore by noise introduced by video tracking procedures. In this paper we propose an algorithm based on dynamic time warping (DTW) as a viable method with which to quantify segmented spatial activity sequences from a video tracking system. DTW is a widely used technique for optimally aligning or warping temporal sequences through minimisation of the distance between their components. The proposed algorithm threshold DTW (TDTW) is capable of accurate spatial sequence distance quantification and is shown using a three class spatial data set to be more robust and accurate than DTW and the discrete hidden markov model (HMM). We also evaluate the application of a band dynamic programming (DP) constraint to TDTW in order to reduce extraneous warping between sequences and to reduce the computation complexity of the approach. Results show that application of a band DP constraint to TDTW improves runtime performance significantly, whilst still maintaining a high precision and recall. 2007 Journal Article http://hdl.handle.net/20.500.11937/7381 http://www.math.ualberta.ca/ijiss/SS-volume-3-07.htm Institute of Scientific Computing and Information restricted
spellingShingle Riedel, Daniel
Venkatesh, Svetha
Liu, Wan-Quan
Treshold dynamic time warping for spatial activity recognition
title Treshold dynamic time warping for spatial activity recognition
title_full Treshold dynamic time warping for spatial activity recognition
title_fullStr Treshold dynamic time warping for spatial activity recognition
title_full_unstemmed Treshold dynamic time warping for spatial activity recognition
title_short Treshold dynamic time warping for spatial activity recognition
title_sort treshold dynamic time warping for spatial activity recognition
url http://www.math.ualberta.ca/ijiss/SS-volume-3-07.htm
http://hdl.handle.net/20.500.11937/7381