Support Vector Machines have recently shown dramatic performance gains in many application areas. We show that the same gains can be realized in the area of speaker recognition via sequence kernels. A sequence kernel provides a numerical comparison of speech utterances as entire sequences rather than a probability at the frame level. We introduce a novel sequence kernel derived from generalized linear discriminants. The kernel has several advantages. First, the kernel uses an explicit expansion into "feature space"--this property allows all of the support vectors to be collapsed into a single vector creating a small speaker model. Second, the kernel retains the computational advantage of generalized linear discriminants trained using mean-squared error training. Finally, the kernel shows dramatic reductions in equal error rates over standard mean-squared error training in matched and mismatched conditions on a NIST speaker recognition task.

A new class of Support Vector Machine (SVM) that is applicable to sequential-pattern recognition such as speech recognition is developed by incorporating an idea of non-linear time alignment into the kernel function. Since the time-alignment operation of sequential pattern is embedded in the new kernel function, standard SVM training and classification algorithms can be employed without further modifications. The proposed SVM (DTAK-SVM) is evaluated in speaker-dependent speech recognition experiments of hand-segmented phoneme recognition. Preliminary experimental results show comparable recognition performance with hidden Markov models (HMMs).

Statistical techniques based on hidden Markov Models (HMMs) with Gaussian emission densities have dominated signal processing and pattern recognition literature for the past 20 years. However, HMMs trained using maximum likelihood techniques suffer from an inability to learn discriminative information and are prone to overfitting and over-parameterization. Recent work in machine learning has focused on models, such as the support vector machine (SVM), that automatically control generalization and parameterization as part of the overall optimization process. In this paper, we show that SVMs provide a significant improvement in performance on a static pattern classification task based on the Deterding vowel data. We also describe an application of SVMs to large vocabulary speech recognition, and demonstrate an improvement in error rate on a continuous alphadigit task (OGI Aphadigits) and a large vocabulary conversational speech task (Switchboard). Issues related to the development and optimization of an SVM/HMM hybrid system are discussed.

Abstract not found

We describe the use of Support Vector Machines for phonetic classification on the TIMIT corpus. Unlike previous work, in which entire phonemes are classified, our system operates in a framewise manner and is intended for use as the front-end of a hybrid system similar to ABBOT. We therefore avoid the problems of classifying variable-length vectors. Our frame-level phone classification accuracy on the complete TIMIT test set is competitive with other results from the literature. In addition, we address the serious problem of scaling Support Vector Machines by using the Kernel Fisher Discriminant. 1.

In this work, motivated by large margin classifiers in machine learning, we propose a novel method to estimate continuous density hidden Markov model (CDHMM) for speech recognition according to the principle of maximizing the minimum muti-class separation margin. The approach is named as large margin HMM. Firstly, we show this type of large margin HMM estimation problem can be formulated as a constrained minimax optimization problem. Secondly, by imposing different constraints to the minimax problem, we propose three solutions to the large margin HMM estimation problem, namely the iterative localized optimization method, the constrained joint optimization method and the semidefinite pro-gramming (SDP) method. These new training methods are evaluated in the isolated E-set recognition task using ISOLET database and the TIDIGITS connected digit string recog-nition task. Experimental results clearly show that the large margin HMMs consistently outperform the conventional HMM training methods. It has been consistently observed that the large margin training method yields significant recognition error rate reduction even on top of some popular discriminative training methods.

In this thesis, Support Vector Machines (SVMs) are applied to the problem of phoneme classification. Given a sequence of acoustic observations and 40 phoneme targets, the task is to classify each observation to one of these targets. Since this task involves multiple classes, one of the main hurdles SVMs must overcome is to extend the inherently binary SVMs to the multi-class case. To do this, several methods are proposed, and their generalisation abilities are measured. It is found that even though some generalisation is lost in the transition, this can still lead to effective classifiers. In addition, a refinement to the SVMs is made to derive estimated posterior probabilities from classifications. Since almost all speech recognition systems are based on statistical models, this is necessary if SVMs are to be used in a full speech recognition system. The best accuracy found was 71.4%, which is competitive with the best results found in literature.

A novel approach for comparing sequences of observations using an explicit-expansion kernel is demonstrated. The kernel is derived using the assumption of the independence of the sequence of observations and a mean-squared error training criterion. The use of an explicit expansion kernel reduces classifier model size and computation dramatically, resulting in model sizes and computation one-hundred times smaller in our application. The explicit expansion also preserves the computational advantages of an earlier architecture based on mean-squared error training.

Traditional speech recognition systems use Gaussian mixture models to obtain the likelihoods of individual phonemes, which are then used as state emission probabilities in hidden Markov models representing the words. In hybrid systems, the Gaussian mixtures are replaced by more discriminant classifiers, leading to an improved performance. Most of the time the classifiers used in such systems are neural networks. Support vector machines have also been used in one-modality audio or visual speech recognition, but never in a multimodal audio-visual system. We propose such a hybrid SVM-HMM speech recognizer, and we show how the multimodal approach leads to better performance than that obtained with any of the two modalities individually. 1.

Speech recognition based on visual information is an emerging research field. We propose here a new system for the recognition of visual speech based on support vector machines which proved to be powerful classifiers in other visual tasks. We use support vector machines to recognize the mouth shape corresponding to di#erent phones produced. To model