TY - GEN
T1 - Using state-level information in the HMM
AU - Li, Hao-Zheng
AU - Liu, Zhi-Qiang
AU - Zhu, Xiang-Hua
N1 - Publication details (e.g. title, author(s), publication statuses and dates) are captured on an “AS IS” and “AS AVAILABLE” basis at the time of record harvesting from the data source. Suggestions for further amendments or supplementary information can be sent to [email protected].
PY - 2004
Y1 - 2004
N2 - In HMM-based pattern recognition, the structure of HMM is predetermined according to some prior knowledge. In the recognition process, we usually make our judgment based on the maximum likelihood of the HMM which unfortunately may lead to incorrect results. In this paper, we analyze the roles of individual hidden states of the HMM and their associated posterior probabilities that reflect the nature of the components in the observation sequence, which should be taken into consideration. For this, we propose to make a full use of the state-level information, e.g., making use of the distribution of the intersection number of state posterior probability trajectories in the recognition process. We apply the proposed methods to phoneme classification on TIMIT speech corpus and show indeed that we are able to achieve about 2% percent improvement in recognition rate over that of the classical HMM.
AB - In HMM-based pattern recognition, the structure of HMM is predetermined according to some prior knowledge. In the recognition process, we usually make our judgment based on the maximum likelihood of the HMM which unfortunately may lead to incorrect results. In this paper, we analyze the roles of individual hidden states of the HMM and their associated posterior probabilities that reflect the nature of the components in the observation sequence, which should be taken into consideration. For this, we propose to make a full use of the state-level information, e.g., making use of the distribution of the intersection number of state posterior probability trajectories in the recognition process. We apply the proposed methods to phoneme classification on TIMIT speech corpus and show indeed that we are able to achieve about 2% percent improvement in recognition rate over that of the classical HMM.
KW - A posteriori state probability
KW - Current state probability
KW - Hidden Markov Models
KW - Learning
KW - Phoneme
KW - Speech recognition
KW - State-level information
KW - Training
UR - https://www.scopus.com/pages/publications/6344249142
UR - https://www.scopus.com/pages/publications/6344249142#tab=impact
U2 - 10.1109/ICMLC.2004.1378574
DO - 10.1109/ICMLC.2004.1378574
M3 - RGC 32 - Refereed conference paper (with host publication)
SN - 0780384032
SN - 9780780384033
VL - 5
T3 - Proceedings of 2004 International Conference on Machine Learning and Cybernetics
SP - 3140
EP - 3145
BT - Proceedings of 2004 International Conference on Machine Learning and Cybernetics
PB - IEEE
T2 - 3rd International Conference on Machine Learning and Cybernetics, ICMLC 2004
Y2 - 26 August 2004 through 29 August 2004
ER -