We propose a modified Hidden Markov Model (HMM) with a view to improve gesture recognition using a moving camera. The conventional HMM is formulated so as to deal with only one feature candidate per frame. However, for a mobile robot, the background and the lighting conditions are always changing, and the feature extraction problem becomes difficult. It is almost impossible to extract a reliable feature vector under such conditions. In this paper, we define a new gesture recognition framework in which multiple candidates of feature vectors are generated with confidence measures and the HMM is extended to deal with these multiple feature vectors. Experimental results comparing the proposed system with feature vectors based on DCT and the method of selecting only one candidate feature point verifies the effectiveness of the proposed technique.
The copyright of the original papers published on this site belongs to IEICE. Unauthorized use of the original or translated papers is prohibited. See IEICE Provisions on Copyright for details.
Copy
Yosuke SATO, Tetsuji OGAWA, Tetsunori KOBAYASHI, "Extension of Hidden Markov Models for Multiple Candidates and Its Application to Gesture Recognition" in IEICE TRANSACTIONS on Information,
vol. E88-D, no. 6, pp. 1239-1247, June 2005, doi: 10.1093/ietisy/e88-d.6.1239.
Abstract: We propose a modified Hidden Markov Model (HMM) with a view to improve gesture recognition using a moving camera. The conventional HMM is formulated so as to deal with only one feature candidate per frame. However, for a mobile robot, the background and the lighting conditions are always changing, and the feature extraction problem becomes difficult. It is almost impossible to extract a reliable feature vector under such conditions. In this paper, we define a new gesture recognition framework in which multiple candidates of feature vectors are generated with confidence measures and the HMM is extended to deal with these multiple feature vectors. Experimental results comparing the proposed system with feature vectors based on DCT and the method of selecting only one candidate feature point verifies the effectiveness of the proposed technique.
URL: https://global.ieice.org/en_transactions/information/10.1093/ietisy/e88-d.6.1239/_p
Copy
@ARTICLE{e88-d_6_1239,
author={Yosuke SATO, Tetsuji OGAWA, Tetsunori KOBAYASHI, },
journal={IEICE TRANSACTIONS on Information},
title={Extension of Hidden Markov Models for Multiple Candidates and Its Application to Gesture Recognition},
year={2005},
volume={E88-D},
number={6},
pages={1239-1247},
abstract={We propose a modified Hidden Markov Model (HMM) with a view to improve gesture recognition using a moving camera. The conventional HMM is formulated so as to deal with only one feature candidate per frame. However, for a mobile robot, the background and the lighting conditions are always changing, and the feature extraction problem becomes difficult. It is almost impossible to extract a reliable feature vector under such conditions. In this paper, we define a new gesture recognition framework in which multiple candidates of feature vectors are generated with confidence measures and the HMM is extended to deal with these multiple feature vectors. Experimental results comparing the proposed system with feature vectors based on DCT and the method of selecting only one candidate feature point verifies the effectiveness of the proposed technique.},
keywords={},
doi={10.1093/ietisy/e88-d.6.1239},
ISSN={},
month={June},}
Copy
TY - JOUR
TI - Extension of Hidden Markov Models for Multiple Candidates and Its Application to Gesture Recognition
T2 - IEICE TRANSACTIONS on Information
SP - 1239
EP - 1247
AU - Yosuke SATO
AU - Tetsuji OGAWA
AU - Tetsunori KOBAYASHI
PY - 2005
DO - 10.1093/ietisy/e88-d.6.1239
JO - IEICE TRANSACTIONS on Information
SN -
VL - E88-D
IS - 6
JA - IEICE TRANSACTIONS on Information
Y1 - June 2005
AB - We propose a modified Hidden Markov Model (HMM) with a view to improve gesture recognition using a moving camera. The conventional HMM is formulated so as to deal with only one feature candidate per frame. However, for a mobile robot, the background and the lighting conditions are always changing, and the feature extraction problem becomes difficult. It is almost impossible to extract a reliable feature vector under such conditions. In this paper, we define a new gesture recognition framework in which multiple candidates of feature vectors are generated with confidence measures and the HMM is extended to deal with these multiple feature vectors. Experimental results comparing the proposed system with feature vectors based on DCT and the method of selecting only one candidate feature point verifies the effectiveness of the proposed technique.
ER -