Summary: | 碩士 === 國立暨南國際大學 === 電機工程學系 === 101 === In human civilization, people gradually increase the demand for technology products, in the past many things in life have to rely on the remote control, keyboard, mouse, input devices and so on. Recent mobile communication, wireless networks, smart phones, and so the technology has become more sophisticated, people and machines to communicate, I believe you can take a more humane, more natural design.
In this thesis, we present two scheme to improve the noise robustness of features in speech recognition. Cepstral mean and variance normalization (CMVN) and cepstral gain normalization (CGN), the processed temporal domain feature sequence is first converted into the modulation spectral domain. The magnitude part of the modulation spectrum is decomposed into overlapped non-uniform sub-band segments, and then each sub-band segment is individually processed by the normalization methods. Recognition experiments implemented on database show that the two methods effectively improve the recognition range of noise environment, like CMVN and CGN, to achieve a more excellent recognition performance.
|