HMM-Based speech recognition using multi-dimensional multi-labeling
作者： M. Nishimura , K. Toshioka
DOI: 10.1109/ICASSP.1987.1169883
关键词: Frame (networking) 、 Multi dimensional 、 Mathematics 、 Sequence labeling 、 Speech recognition 、 Vector quantization 、 Word recognition 、 Word error rate 、 Task (computing) 、 Hidden Markov model 、 Artificial intelligence 、 Pattern recognition
摘要: This paper describes a new vector quantization (VQ; so-called labeling) method of speech recognition system based on hidden Markov model (HMM). For improving the VQ accuracy in simple manner, "multi-labeling" which generates multiple labels at each frame was introduced while keeping conventional HMM formulation. Furthermore, order to represent characteristics accurately and effectively, "multi-dimensional labeling" also quantizes features such as spectral dynamics spectrum independently. labeling tested an isolated word task using 150 Japanese confusable words. The error rate roughly reduced 1/2 or less compared with method.
sci-hub.se 参考文章(8)
S. Soudoplatoff, Markov modeling of continuous parameters in speech recognition international conference on acoustics, speech, and signal processing. ,vol. 11, pp. 45- 48 ,(1986) , 10.1109/ICASSP.1986.1169180
H. Bourlard, C. Wellekens, H. Ney, Connected digit recognition using vector quantization international conference on acoustics, speech, and signal processing. ,vol. 9, pp. 413- 416 ,(1984) , 10.1109/ICASSP.1984.1172585
Frederick Jelinek, Continuous speech recognition by statistical methods Proceedings of the IEEE. ,vol. 64, pp. 532- 556 ,(1976) , 10.1109/PROC.1976.10159
L. R. Rabiner, M. M. Sondhi, S. E. Levinson, A Vector Quantizer Combining Energy and LPC Parameters and Its Application to Isolated Word Recognition AT&T Bell Laboratories Technical Journal. ,vol. 63, pp. 721- 735 ,(1984) , 10.1002/J.1538-7305.1984.TB00104.X
Teruo Okuda, Eiichi Tanaka, Tamotsu Kasai, A Method for the Correction of Garbled Words Based on the Levenshtein Metric IEEE Transactions on Computers. ,vol. 25, pp. 172- 178 ,(1976) , 10.1109/TC.1976.5009232
S. Furui, Speaker-independent isolated word recognition based on emphasized spectral dynamics international conference on acoustics, speech, and signal processing. ,vol. 11, pp. 1991- 1994 ,(1986) , 10.1109/ICASSP.1986.1168654
B. Juang, L. Rabiner, S. Levinson, M. Sondhi, Recent developments in the application of hidden Markov models to speaker-independent isolated word recognition international conference on acoustics, speech, and signal processing. ,vol. 10, pp. 9- 12 ,(1985) , 10.1109/ICASSP.1985.1168453
K. Sugawara, M. Nishimura, K. Toshioka, M. Okochi, T. Kaneko, Isolated word recognition using hidden Markov models international conference on acoustics, speech, and signal processing. ,vol. 10, pp. 1- 4 ,(1985) , 10.1109/ICASSP.1985.1168452