Signal Processing for Robust Speech Recognition

SIGNAL PROCESSING FOR ROBUST SPEECH RECOGNITION Fu-Hua Liu, Pedro J. Moreno, Richard M. Stem, Alejandro Acero Department of Electrical and Computer Engineering School of Computer Science Carnegie Mellon University Pittsburgh, PA 15213 ABSTRACT This paper describes a series of ceps~al-based compensation pro- cedures that render the SPHINX-II system more robust with respect to acoustical environment. The first algorithm, phone- dependent cepstral compensation, is similar in concept to the pre- viously-described MFCDCN method, except that cepstral com- pensation vectors are selected according to the current phonetic hypothesis, rather than on the basis of SNR or VQ codeword iden- tity. We also describe two procedures to accomplish adaptation of the VQ codebook for new environments, as well as the use of reduced-bandwidth f~equency analysis to process telephone-band- width speech. Use of the various compensation algorithms in con- sort produces a reduction of error rates for SPHINX-II by as much as 40 percent relative to the rate achieved with eepstral mean nor- realization alone, in both development test sets and in the context of the 1993 ARPA CSR evaluations. 1. INTRODUCTION A continuing problem with current speech recognition technology is that of lack of robustness with respect to environmental variabil- ity. For example, the use of microphones other than the ARPA standard Sennheiser HM--414 "close-talking" headset (CLSTLK) severely degrades the performance of systems like the original SPHINX system, even in a relatively quiet office environment [e.g. 1,2]. Applications such as speech recognition in automobiles, over telephones, on a factory floor, or outdoors demand an even greater degree of environmental robustness. In this paper we describe and compare the performance of a series of cepstrum-based procedures that enable the CMU SPHINX-II [8] speech recognition system to maintain a high level of recogni- tion accuracy over a wide variety of acoustical environments. We also discuss the aspects of these algorithms that appear to have contributed most significantly to the success of the SPHINX-II system in the 1993 ARPA CSR evaluations for microphone inde- pendence (Spoke 5) and calibrated noise sources (Spoke 8). In previous years we described the performance of cepstral map- ping procedures such as the CDCN algorithm, which is effective but fairly computationally costly [2]. More recently we discussed the use of eepstral highpass-filtering algorithms [such as the popu- lax RASTA and cepstral-mean-normalization algorithms (CMN) [6]. These algorithms are very simple to implement but somewhat limited in effectiveness, and CMN is now part of baseline process- ing for the CMU and many other systems. In this paper we describe several new procedures that when used in consort can provide as much as an additional 40 percent improvement over baseline processing with CMN. These tech- niques include: • Phone-dependent cepstral compensation • Environmental interpolation of compensation vectors • Codebook adaptation • Reduced-band analysis for telephone-bandwidth speech. • Silence codebook adaptation In Sec. 2 we describe these compensation procedures in detail, and we examine their effect on recognition accuracy in Secs. 3 and 4. 2. ENVIRONMENTAL COMPENSATION ALGORITHMS We begin this section by reviewing the …

Content truncated to 500 words