Paper Abstract and Keywords |
Presentation |
2006-09-29 14:30
Research concerned with speech recognition by using wavelet transform Cao Yue, Toshitaka Nagano, Hidehiko Arai, Yuya Usami, Kazushi Takahashi, Kenji Kudo, Masatoshi Sekine (TUAT) |
Abstract |
(in Japanese) |
(See Japanese page) |
(in English) |
The method of settlement that recognizes the voice is existing variously now. The wavelet transform is has an excellent ability in the fixed non-length signal, and an algorithm efficient compared with the break-up signal. This research proposes the design of the aural model who uses the multiple resolution expression of the break-up waving let conversion by the Haar function as a preprocessing of the speech-recognition system. The simulation results of the waving let conversion are compared by two design methods, and the advantage of each method brings it together. Moreover, it goes to FPGA on hwModele that is the PCI device developed in this laboratory to mount. |
Keyword |
(in Japanese) |
(See Japanese page) |
(in English) |
Voice Recognition / Wavelet Transform / Muti-Resolution / Aural Model / / / / |
Reference Info. |
IEICE Tech. Rep., vol. 106, no. 267, EA2006-51, pp. 19-24, Sept. 2006. |
Paper # |
EA2006-51 |
Date of Issue |
2006-09-22 (EA) |
ISSN |
Print edition: ISSN 0913-5685 |
Download PDF |
|