NOISE ROBUST SPEECH RECOGNITION USING SUBBAND-CROSS-CORRELATION ANALYSIS

This paper describes subband-crosscorrelation analysis (SBXCOR) using two input channel signals. SBXCOR is an extended signal processing tec hnique of subband-autocorrelation analysis (SBCOR) that extracts perio dicities associated with the inverse of center frequencies present in speech signals. In addition, to extract more periodicity information a ssociated with the inverse of center frequencies, the multi-delay weig hting (MDW) processing is applied to SBXCOR. In experiments, the noise robustness of SBXCOR is evaluated using a DTW word recognizer under ( 1) a simulated acoustic condition with white noise and (2) a real acou stic condition in a sound proof room with human speech-like noise. As the results, under the simulated acoustic condition, it is shown that SBXCOR is more robust than the conventional one-channel SBCOR, but les s robust than SBCOR extracted From the two-channel-summed signal. Furt hermore, by applying MDW processing, the performance of SBXCOR improve d about 2% at SNR 0 dB. The resultant performance of SBXCOR with MDW p rocessing was much better than those of smoothed group delay spectrum (SGDS) and mel-filterbank cepstral coefficient (MFCC) below SNR 10 dB. The results under the real acoustic condition were almost the same as the simulated acoustic condition.