This paper describes the application of a discriminative HMM parameter estimation technique called Frame Discrimination (FD), to medium and large vocabulary continuous speech recognition. Previous work has shown that FD training can give better results than maximum mutual information (MMI) training for small tasks. The use of FD for much larger tasks required the development of a technique to be able to rapidly find the most likely set of Gaussians for each frame in the system. Experiments on the Resource Management and North American Business tasks show that FD training can give comparable improvements to MMI, but is less computationally intensive. 1. INTRODUCTION Previous research has shown that the accuracy of a speech recognition system trained using Maximum Likelihood Estimation (MLE) can often be improved further using discriminative training. All such techniques normally give much greater improvements in recognition accuracy on the training data than on the test set except wh...

This work presents experiments to recognize pattern sequences using hidden Markov models (HMMs). The pattern sequences in the experiments are computed from speech signals and the recognition task is to decode the corresponding phoneme sequences. The training of the HMMs of the phonemes using the collected speech samples is a difficult task because of the natural variation in the speech. Two neural computing paradigms, the Self-Organizing Map (SOM) and the Learning Vector Quantization (LVQ) are used in the experiments to improve the recognition performance of the models. A HMM consists of sequential states which are trained to model the feature changes in the signal produced during the modeled process. The output densities applied in this work are mixtures of Gaussian density functions. SOMs are applied to initialize and train the mixtures to give a smooth and faithful presentation of the feature vector space defined by the corresponding training samples. The SOM maps similar feature vect...

The Hidden Markov Model (HMMs) is one of the most successful modeling approaches for acoustic events in speech recognition, and more recently it has proven useful for several problems in biological sequence analysis. Although the HMM is good at capturing the temporal nature of processes such as speech, it has a very limited capacity for recognizing complex patterns involving more than first order dependencies in the observed data sequences. This is due to the first order state process and the assumption of state conditional independence between observations. Artificial Neural Networks (NNs) are almost the opposite: they cannot model dynamic, temporally extended phenomena very well, but are good at static classification and regression tasks. Combining the two frameworks in a sensible way can therefore lead to a more powerful model with better classification abilities. The overall aim of this work has been to develop a probabilistic hybrid of hidden Markov models and neural networks and ...

In this paper we compare three frameworks for discriminative training of continuous-density hidden Markov models (CD-HMMs). Specifically, we compare two popular frameworks, based on conditional maximum likelihood (CML) and minimum classification error (MCE), to a new framework based on margin maximization. Unlike CML and MCE, our formulation of large margin training explicitly penalizes incorrect decodings by an amount proportional to the number of mislabeled hidden states. It also leads to a convex optimization over the parameter space of CD-HMMs, thus avoiding the problem of spurious local minima. We used discriminatively trained CD-HMMs from all three frameworks to build phonetic recognizers on the TIMIT speech corpus. The different recognizers employed exactly the same acoustic front end and hidden state space, thus enabling us to isolate the effect of different cost functions, parameterizations, and numerical optimizations. Experimentally, we find that our framework for large margin training yields significantly lower error rates than both CML and MCE training. Index Terms — speech recognition, discriminative training, MMI, MCE, large margin, phoneme recognition 1.

In this paper we introduce a set of related confidence measures for large vocabulary continuous speech recognition (LVCSR) based on local phone posterior probability estimates output by an acceptor HMM acoustic model. In addition to their computational efficiency, these confidence measures are attractive as they may be applied at the state-, phone-, word- or utterance-levels, potentially enabling discrimination between different causes of low confidence recognizer output, such as unclear acoustics or mismatched pronunciation models. We have evaluated these confidence measures for utterance verification using a number of different metrics. Experiments reveal several trends in `profitability of rejection', as measured by the unconditional error rate of a hypothesis test. These trends suggest that crude pronunciation models can mask the relatively subtle reductions in confidence caused by out-of-vocabulary (OOV) words and disfluencies, but not the gross model mismatches elicited by non-sp...

Many conventional speech recognition systems are based on the use of hidden Markov models (HMM) within the context of discriminant-based pattern classification. While the speech recognition objective is a low rate of misclassification, HMM design has been traditionally approached via maximum likelihood (ML) modeling which is, in general, mismatched with the minimum error objective and hence suboptimal. Direct minimization of the error rate is difficult because of the complex nature of the cost surface, and has only been addressed recently by discriminative design methods such as generalized probabilistic descent (GPD). While existing discriminative methods offer significant benefits, they commonly rely on local optimization via gradient descent whose performance suffers from the prevalence of shallow local minima. As an alternative, we propose the deterministic annealing (DA) design method that directly minimizes the error rate while avoiding many poor local minima of the cost. DA is derived from fundamental principles of statistical physics and information theory. In DA, the HMM classifier's decision is randomized and its expected error rate is minimized subject to a constraint on the level of randomness which is measured by the Shannon entropy. The entropy constraint is gradually relaxed, leading in the limit of zero entropy to the design of regular nonrandom HMM classifiers. An efficient forward--backward algorithm is proposed for the DA method. Experiments on synthetic data and on a simplified recognizer for isolated English letters demonstrate that the DA design method can improve recognition error rates over both ML and GPD methods.

Recent advances in automatic speech recognition are accomplished by designing a plug-in maximum a posteriori decision rule such that the forms of the acoustic and language model distributions are specified and the parameters of the assumed distributions are estimated from a collection of speech and language training corpora. Maximum-likelihood point estimation is by far the most prevailing training method. However, due to the problems of unknown speech distributions, sparse training data, high spectral and temporal variabilities in speech, and possible mismatch between training and testing conditions, a dynamic training strategy is needed. To cope with the changing speakers and speaking conditions in real operational conditions for high-performance speech recognition, such paradigms incorporate a small amount of speaker and environment specific adaptation data into the training process. Bayesian adaptive learning is an optimal way to combine

Recently there has been significant interest in developing new acoustic models for speech recognition. One such model, that allows complex dependencies to be represented, is the augmented statistical model. This incorporates additional dependencies by constructing a local exponential expansion of a standard HMM. Unfortunately, the resulting model often has an intractable normalisation term, rendering training difficult for all but binary classification tasks. In this paper, conditional augmented (C-Aug) models are proposed as an attractive alternative. Instead of modelling utterance likelihoods and inferring decision boundaries, C-Aug models directly model the posterior probability of class labels, conditioned on the utterance. The resulting model is easy to normalise and can be trained using conditional maximum likelihood estimation. In addition, as a convex model, the optimisation converges to a global maximum. 1.

This paper studies the problem of ergodicity of transition probabilitymatricesinMarkovian models, such as hidden Markov models (HMMs), and how itmakes very di cult the task of learning to represent long-term context for sequential data. This phenomenon hurts the forward propagation of long-term context information, as well as learning a hidden state representation to represent long-term context, which depends on propagating credit information backwards in time. Using results from Markov chain theory, weshow that this problem of di usion of context and credit is reduced when the transition probabilities approach 0 or 1, i.e., the transition probability matrices are sparse and the model essentially deterministic. The results found in this paper apply to learning approachesbasedon continuous optimization, such asgradient descent and the Baum-Welch algorithm. 1.

Abstract. This paper describes a clustering methodology for temporal data using hidden Markov model(HMM) representation. The proposed method improves upon existing HMM based clustering methods in two ways: (i) it enables HMMs to dynamically change its model structure to obtain a better t model for data during clustering process, and (ii) it provides objective criterion function to automatically select the clustering partition. The algorithm is presented in terms of four nested levels of searches: (i) the search for the number of clusters in a partition, (ii) the search for the structure for a xed sized partition, (iii) the search for the HMM structure for each cluster, and (iv) the search for the parameter values for each HMM. Preliminary experiments with arti cially generated data demonstrate the e ectiveness of the proposed methodology. 1