ARPA Wall Street Journal Research Articles

Fast likelihood computation techniques in nearest-neighbor based search for continuous speech recognition

This paper describes two effective algorithms that reduce the computational complexity of state likelihood computation in mixture-based Gaussian speech recognition systems. We consider a baseline recognition system that uses nearest-neighbor search and partial distance elimination (PDE) to compute state likelihoods. The first algorithm exploits the high dependence exhibited among subsequent feature vectors to predict the best scoring mixture for each state. The method, termed best mixture prediction (BMP), leads to further speed improvement in the PDE technique. The second technique, termed feature component reordering (FCR), takes advantage of the variable contribution levels made to the final distortion score for each dimension of the feature and mean space vectors. The combination of two techniques with PDE reduces the computational time for likelihood computation by 29.8% over baseline likelihood computation. The algorithms are shown to yield the same accuracy level without further memory requirements for the November 1992 ARPA Wall Street Journal (WSJ) task.

Comparison of discriminative training criteria and optimization methods for speech recognition

The aim of this work is to build up a common framework for a class of discriminative training criteria and optimization methods for continuous speech recognition. A unified discriminative criterion based on likelihood ratios of correct and competing models with optional smoothing is presented. The unified criterion leads to particular criteria through the choice of competing word sequences and the choice of smoothing. Analytic and experimental comparisons are presented for both the maximum mutual information (MMI) and the minimum classification error (MCE) criterion together with the optimization methods gradient descent (GD) and extended Baum (EB) algorithm. A tree search-based restricted recognition method using word graphs is presented, so as to reduce the computational complexity of large vocabulary discriminative training. Moreover, for MCE training, a method using word graphs for efficient calculation of discriminative statistics is introduced. Experiments were performed for continuous speech recognition using the ARPA wall street journal (WSJ) corpus with a vocabulary of 5k words and for the recognition of continuously spoken digit strings using both the TI digit string corpus for American English digits, and the SieTill corpus for telephone line recorded German digits. For the MMI criterion, neither analytical nor experimental results do indicate significant differences between EB and GD optimization. For acoustic models of low complexity, MCE training gave significantly better results than MMI training. The recognition results for large vocabulary MMI training on the WSJ corpus show a significant dependence on the context length of the language model used for training. Best results were obtained using a unigram language model for MMI training. No significant correlation has been observed between the language models chosen for training and recognition.

Compensation for speech recognition in degraded acoustical environments

The accuracy of speech recognition systems degrades when operated in adverse acoustical environments. This paper discusses two ways in which more detailed mathematical descriptions of the effects of environmental degradation can improve speech recognition accuracy using both ‘‘data-driven’’ and ‘‘model-based’’ compensation strategies. Data-driven methods learn environmental characteristics through direct comparisons of speech recorded in the noisy environment with the same speech recorded under optimal conditions. Model-based methods use a mathematical model of the environment and attempt to use samples of the degraded speech to estimate model parameters. Two approaches to data-driven compensation, RATZ and STAR, are described, as well as a new approach to model-based compensation, referred to as the vector Taylor series (VTS) algorithm. Compensation algorithms are evaluated in a series of experiments measuring recognition accuracy for speech from the ARPA Wall Street Journal database that is corrupted by artificially added noise at various signal-to-noise ratios (SNRs). For any particular SNR, the greatest recognition accuracy obtained using a practical compensation algorithm is observed when that system is trained using noisy data at that SNR. The RATZ, VTS, and STAR algorithms achieve this bound at global SNRs as low as 15, 10, and 5 dB, respectively. [Work supported by ARPA.]

Genones: generalized mixture tying in continuous hidden Markov model-based speech recognizers

An algorithm is proposed that achieves a good tradeoff between modeling resolution and robustness by using a new, general scheme for tying of mixture components in continuous mixture-density hidden Markov model (HMM)-based speech recognizers. The sets of HMM states that share the same mixture components are determined automatically using agglomerative clustering techniques. Experimental results on ARPA's Wall Street Journal corpus show that this scheme reduces errors by 25% over typical tied-mixture systems. New fast algorithms for computing Gaussian likelihoods-the-most time-consuming aspect of continuous-density HMM systems-are also presented. These new algorithms-significantly reduce the number of Gaussian densities that are evaluated with little or no impact on speech recognition accuracy.