Automated recognition of sequential patterns in captured motion streams

Authors:
Liqun Deng;Howard Leung;Naijie Gu;Yang Yang
Affiliations:
University of Science and Technology of China, Hefei, China and City University of Hong Kong, Hong Kong;City University of Hong Kong, Hong Kong;University of Science and Technology of China, Hefei, China;University of Science and Technology of China, Hefei, China and City University of Hong Kong, Hong Kong
Venue:
WAIM'10 Proceedings of the 11th international conference on Web-age information management
Year:
2010

Citing 5
Cited 1

Simultaneous gesture segmentation and recognition based on forward spotting accumulative HMMs

Pattern Recognition
Matching incomplete time series with dynamic time warping: an algorithm and an application to post-stroke rehabilitation

Artificial Intelligence in Medicine
APCAS: an approximate approach to adaptively segment time series stream

APWeb/WAIM'07 Proceedings of the joint 9th Asia-Pacific web and 8th international conference on web-age information management conference on Advances in data and web management
Human motion recognition using Isomap and dynamic time warping

Proceedings of the 2nd conference on Human motion: understanding, modeling, capture and animation
Gesture spotting in continuous whole body action sequences using discrete hidden markov models

GW'05 Proceedings of the 6th international conference on Gesture in Human-Computer Interaction and Simulation

Generalized Model-Based Human Motion Recognition with Body Partition Index Maps

Computer Graphics Forum

Quantified Score

Hi-index	0.00

Visualization

Abstract

Motion capture data has been frequently used in computer animation and video games. Motions are often captured in a continuous manner such that a motion contains multiple patterns joined sequentially without obvious breakpoints between them. It is challenging to learn the captured motions as it requires both segmentation and recognition. In this paper, a new method based on an extension of open-end dynamic time warping (OE-DTW) is proposed to automatically segment and recognize sequential patterns in motion streams. To enhance the performance, we introduce a global constraint of K-Repetition on OE-DTW and a flexible end point detection scheme. In the experiments, we applied our method on different classes of dance motions and demonstrated the effectiveness of our method by comparing with existing approach.