http://chineseinput.net/에서 pinyin(병음)방식으로 중국어를 변환할 수 있습니다.
변환된 중국어를 복사하여 사용하시면 됩니다.
허만탁 釜山工業大學校 1992 論文集 Vol.34 No.-
In this study, the first and the second formants of Korean 6 single vowels over chirdern through the olds and, male and female are extracted. The extraction method is discribed. The young's voice was studied with deal. The distribution of formants was not very clear, particulaly about/a/, /oe/, /o/ and /u/. This might caused by similar tong's position and lip's shape. Especially, /o/ and /u/ are difficult to seperate in two. Beside, old's and children's voise were studied, too. Their feature was different from young's. sampling rate is 10[KHz] and resolution is 12 bit.
권영욱,허만탁 釜慶大學校 1997 釜慶大學校 論文集 Vol.2 No.2
Speech recognition in noise environments requires a good estimate of the noise power spectrum in the preprocessing technique In this paper, we em-ploy the mixture Gaussian technique for the estimation of moise spectrum This technique has advantage over ther noise estimation methods in which it does not require speech/non-speech detection and can estimate slowly-varying noise spectra. According to the speaker-independent isolated word rec-ognition in both colored Gaussian noise environments under various SNR conditions, mixture Gaussian based spectral subrtaction method yields superier performance to the one with conventional noise estimation method using the spectral acerge of initial frames during on-speech period and histogram-technique-based spectral subtraction method.
許萬鐸 釜山工業大學校 1995 論文集 Vol.37 No.-
In this paper, a diffusion filter bank which needs very less hardware than conventional one is introduced, and the results of extracting spectum from real speech signals by use of that are presented. The spectra extracted from the signals by simmulation have so smooth envelope that we can estimate it is suitable for speech recognition which demand many channels with diffrent center freqencies, and smooth spectrum of speech signals.