Noise reduction for automatic speech recognition.

摘要

Noise reduction procedure for automatic speech recognition, comprising: - computing a magnitude spectrum | Yk (m) | a noisy speech containing a clean speech to be recognized and noise affecting the clean speech; - calculating a power spectrum (| Yk (m) | 2) of the noisy speech; - calculate an estimate (| Xk (m) | 2) of a power spectrum of the clean speech; - computing an estimate (| Dk (m) | 2) of a power spectrum of the noise; - calculate an estimate (Ek (m)) of a signal-to-noise ratio a priori as a function of the estimate (| Xk (m) | 2) of the power spectrum of the clean speech and the estimate (| Dk (m) | 2) of the power spectrum of the noise; - computing an estimate (Ek (m)) of a signal-to-noise post as a function of the power spectrum (| Yk (m) | 2) of the noisy speech and the estimate (| Dk (m) | 2) power spectrum of the noise; - computing an attenuation gain (Gk (m)) as a function of the estimate (Ek (m)) of the signal to noise priori estimation (k (m)) of the signal to noise posteriori; - calculate an estimate (| Xk (m) |) of a magnitude spectrum of the clean speech as a function of the magnitude spectrum (| Yk (m) |) of the noisy speech and the attenuation gain (Gk (m)); characterized in that calculating estimates (Ek (m), Yk (m)) of the signal-to-noise priori and a posteriori comprising: