Robust Speech Recognition Using a Harmonic Model
许超, 曹志刚 · Acta Scientiarum Naturalium Universitatis Sunyatseni · 2004
Automatic speech recognition under conditions of a noisy environment remains a challenging problem. Traditionally, methods focused on noise structure, such as spectral subtraction, have been employed to address this problem, and thus the performance of such methods depends on the accuracy in noise estimation. In this paper, an alternative method, using a harmonic-based spectral reconstruction algorithm, is proposed for the enhancement of robust automatic speech recognition. Neither noise estimation nor noise-model training are required in the proposed approach. A spectral subtraction integrated autocorrelation function is proposed to determine the pitch for the harmonic model. Recognition results show that the harmonic-based spectral reconstruction approach outperforms spectral subtraction in the middle- and lowsignal noise ratio (SNR) ranges. The advantage of the proposed method is more manifest for non-stationary noise, as the algorithm does not require an assumption of stationary noise.