is granted. A Maximum-Likelihood Approach to Stochastic Matching for Robust Speech Recognition Ananth Sankar 2 and Chin-Hui Lee Speech Research Department AT&T Bell Laboratories Murray Hill, NJ 07974 1 Introduction Recently there has been much interest in the problem of improving the performance of automatic speech recognition (ASR) systems in adverse environments. When there is a mismatch between the training and testing environments, ASR systems suffer a degradation in performance. The goal of robust speech recognition is to remove the effect of this mismatch so as to bring the recognition performance as close as possible to the matched conditions. In speech recognition, the speech is usually modeled by a set of hidden Markov models (HMM) X . During recognition the observed utterance Y is decoded using these models. Due to the mismatch between training and testing conditions, this often results in a degradation in performance compared to the matched conditions. The mismatch b...

This paper addresses the problem of automatic speech recognition in the presence of interfering noise. It focuses on the Parallel Model Combination (PMC) scheme, which has been shown to be a powerful technique for achieving noise robustness. Most experiments reported on PMC to date have been on small, 10-50 word vocabulary systems. Experiments on the Resource Management (RM) database, a 1000 word continuous speech recognition task, reveal compensation requirements not highlighted by the smaller vocabulary tasks. In particular, that it is necessary to compensate the dynamic parameters as well as the static parameters to achieve good recognition performance. The database used for these experiments was the RM speaker independent task with either Lynx Helicopter noise or Operation Room noise from the NOISEX-92 database added. The experiments reported here used the HTK RM recogniser developed at CUED modified to include PMC based compensation for the static, delta and delta-delta parameters. After training on clean speech data,the performance of the recogniser was found to be severely degraded when noise was added to the speech signal at between 10dB and 18dB. However, using PMC the performance was restored to a level comparable with that obtained when training directly in the noise corrupted environment. 1

rely almost exclusively on the acoustic speech signal and, consequently, these systems often perform poorly in noisy environments [I]. Attempts to clean up the acoustic input have had limited success [2]. Another approach is to use other sources of speech information, such as visual speech signals. The perception of acoustic speech by humans can be affected by the visible speech signals [3-51. Specifically, when the acoustic signal is degraded by noise, the visual signal can provide supplemental speech information that improves speech perception [6-81. When no acoustic signal is available, as for the profoundly deaf, the visual signal alone can provide speech information through lip reading [9- 1 I]. Here we answer two questions: Can the speech information conveyed by visual speech signals be extracted automatically? How can this information be combined with information from the acoustic signal to improve automat

The paper begins by discussing the difficulties in obtaining repeatable results in speech recognition. Theoretical arguments are presented for and against copying human auditory properties in automatic speech recognition. The "standard" acoustic analysis for automatic speech recognition, consisting of melscale cepstrum coefficients and their temporal derivatives, is described. Some variations and extensions of the standard analysis --- PLP, cepstrum correlation methods, LDA, and variants on log power --- are then discussed. These techniques pass the test of having been found useful at multiple sites, especially with noisy speech. The extent to which auditory properties can account for the advantage found for particular techniques is considered. It is concluded that the advantages do not in fact stem from auditory properties, and that there is so far little or no evidence that the study of the human auditory system has contributed to advances in automatic speech recognition. Contributio...

For practical applications speech recognition systems need to be insensitive to differences between training and test acoustic conditions. Differences in the acoustic environment may result from various sources, such as ambient background noise, channel variations and speaker stress. These differences can dramatically degrade the performance of a speech recognition system. A wide range of techniques have been proposed for achieving noise robustness. This paper considers one particular approach to model-based compensation, predictive model-based compensation, which has been shown to achieve good noise robustness in a wide range of acoustic environments. The characteristic of these schemes is that they combine a speech model with an additive noise model, a channel model and, in the general case, a speaker stress model, to generate a corrupted-speech model. The general theory of these predictive techniques is discussed. Various approximations for rapidly performing the model combination stage have been proposed and are reviewed in this paper. The advantages and the limitations of such a predictive approach to noise robustness are also discussed. In addition, methods for combining predictive schemes with schemes which make use of speech data in the new environment, adaptive schemes, are detailed. This combined approach overcomes some of the limitations of the predictive schemes. 1 The author is now at the IBM T.J. Watson Research Center, Yorktown Heights, NY 10598, USA. 1

Perceptual harmonic cepstral coefficients (PHCC) are proposed as features to extract for speech recognition. Pitch estimation and classification into voiced, unvoiced, and transitional speech are performed by a spectro-temporal auto-correlation technique. A peak picking algorithm is then employed to precisely locate pitch harmonics. A weighting function, which depends on the classification and the pitch harmonics, is applied to the power spectrum and ensures accurate representation of the voiced speech spectral envelope. The harmonics weighted power spectrum undergoes mel-scaled band-pass filtering, and the logenergy of the filters' output is discrete cosine transformed to produce cepstral coefficients. For perceptual considerations, within-filter cubic-root amplitude compression is applied to reduce amplitude variation without compromise of the gain invariance properties. Experiments show substantial recognition gains of PHCC over MFCC, with 48% and 15% error rate reduction for the Mandarin digit database and E-set, respectively.

this paper.

Model compensation is a standard way of improving speech recognisers’ robustness to noise. Most model compensation techniques produce diagonal covariances. However, this fails to handle any changes in the feature correlations due to the noise. This paper presents a scheme that allows full-covariance matrices to be estimated. One problem is that full covariance matrix estimation will be more sensitive approximations, those for the dynamic parameters are known to crude. In this paper a linear transformation of a window of consecutive frames is used as the basis for dynamic parameter compensation. A second problem is that the resulting full covariance matrices slow down decoding. This is addressed by using predictive linear transforms that decorrelate the feature space, so that the decoder can then use diagonal covariance matrices. On a noise-corrupted Resource Management task, the proposed scheme outperformed the standard VTS compensation scheme.

Theoretical and practical issues of some of the problems in robust automatic speech recognition (ASR) and some of the techniques that address them are presented in this report. The problem of the robustness of the ASR in real--life (as opposed to laboratory) conditions is paramount to the widespread deployment of speech enabled products. The report reviews techniques used so far for robust ASR, ranging from simple spectrum subtraction to various types of model adaptation. A possible connection of robust ASR with the computational auditory scene analysis (CASA), methods for local Signal--to--Noise Ratio (SNR) estimation and classification/scoring with on--line adapted statistical models is discussed. The main focus is on the techniques that would allow for incorporation of CASA and local SNR estimates (used as methods for speech/non--speech separation) into the present prevailing stochastic pattern matching paradigms -- Hidden Markov models (HMM) and artificial neural networks (ANN). Th...

In real speaker verification applications, additive or convolutive noise creates a mismatch between training and recognition environments, degrading performance. Parallel Model Combination (PMC) is used successfully to improve the noise robustness of Hidden Markov Model (HMM) based speech recognisers [5]. This paper presents the results of applying PMC to compensate for additive noise in HMM-based text-dependent speaker verification. Speech and noise data were obtained from the YOHO [6] and NOISEX-92 databases [13] respectively. Speaker recognition Equal Error Rates (EER) are presented for noise-contaminated speech at different signal-to-noise ratios (SNRs) and different noise sources. For example, average EER for speech in operations room noise at 6dB SNR dropped from approximately 20% un-compensated to less than 5% using PMC. Finally, it is shown that speaker recognition performance is relatively insensitive to the exact value of the parameter that determines the relative amplitudes of the speech and noise components of the PMC model.