Traditional Hidden Markov Model Research Articles

Face recognition based on multi-class mapping of Fisher scores

A new hidden Markov model (HMM) based feature generation scheme is proposed for face recognition (FR) in this paper. In this scheme, HMM method is used to model classes of face images. A set of Fisher scores is calculated through partial derivative analysis of the parameters estimated in each HMM. These Fisher scores are further combined with some traditional features such as log-likelihood and appearance based features to form feature vectors that exploit the strengths of both local and holistic features of human face. Linear discriminant analysis (LDA) is then applied to analyze these feature vectors for FR. Performance improvements are observed over stand-alone HMM method and Fisher face method which uses appearance based feature vectors. A further study reveals that, by reducing the number of models involved in the training and testing stages of LDA, the proposed feature generation scheme can maintain very high discriminative power at much lower computational complexity comparing to the traditional HMM based FR system. Experimental results on a public available face database are provided to demonstrate the viability of this scheme.

Recognition of Visual Speech Elements Using Adaptively Boosted Hidden Markov Models

The performance of automatic speech recognition (ASR) system can be significantly enhanced with additional information from visual speech elements such as the movement of lips, tongue, and teeth, especially under noisy environment. In this paper, a novel approach for recognition of visual speech elements is presented. The approach makes use of adaptive boosting (AdaBoost) and hidden Markov models (HMMs) to build an AdaBoost-HMM classifier. The composite HMMs of the AdaBoost-HMM classifier are trained to cover different groups of training samples using the AdaBoost technique and the biased Baum-Welch training method. By combining the decisions of the component classifiers of the composite HMMs according to a novel probability synthesis rule, a more complex decision boundary is formulated than using the single HMM classifier. The method is applied to the recognition of the basic visual speech elements. Experimental results show that the AdaBoost-HMM classifier outperforms the traditional HMM classifier in accuracy, especially for visemes extracted from contexts.

Tone Modeling for Continuous Mandarin Speech Recognition

Tone study is very important for Mandarin speech recognition. In this paper, a Mixture Stochastic Polynomial Tone Model (MSPTM) is proposed for tone modeling in continuous Mandarin speech. In this model the pitch contour, main representative of tone pattern, is described as a mixed stochastic trajectory. The mean trajectory is represented by a polynomial function of normalized time while the variance is time varying. Effective training and tone recognition algorithms were developed. The experimental results based on the proposed MSPTM showed 40.7% tone recognition error rate reduction relative to the traditional Hidden Markov Model (HMM) tone model. We also present a decision tree based approach to learning the tone pattern variation in continuous speech. The phonetic and linguistic factors that may affect the tone patterns were taken into consideration while constructing the tree. After the tree was established, 28 different tone patterns were obtained. We found that in addition to the tone of the neighboring syllable, Consonant/Vowel type of the syllable and the position of the syllable in the utterance also made important contributions to tone pattern variations in continuous speech. Finally, a new approach of integrating tone information into the search process at word level is discussed. Experiments on continuous Mandarin speech recognition showed that the new tone model and tone information integration method were efficient, achieving a 16.2% relative character error rate reduction.

The Hidden Markov Model of co-articulation and its application to the continuous speech recognition

The co-articulation is one of the main reasons that makes the speech recognition difficult. However, the traditional Hidden Markov Models(HMM) can not model the co-articulation, because they depend on the first-order assumption. In this paper, for modeling the co-articulation, a more perfect HMM than traditional first order HMM is proposed on the basis of the authors’ previous works(1997, 1998) and they give a method in that this HMM is used in continuous speech recognition by means of multilayer perceptrons(MLP), i.e. the hybrid HMM/MLP method with triple MLP structure. The experimental result shows that this new hybrid HMM/MLP method decreases error rate in comparison with authors’ previous works.

Recognition of on-line cursive Korean characters combining statistical and structural methods

In this paper, we propose a hybrid recognition method, and show its usefulness in recognizing online cursive Korean characters. A finite state network is constructed to represent the rules of character composition from graphemes. In the network, each arc and node expands into statistical and structural recognizers, respectively. The statistical recognizer produces intermediate recognition results in traditional hidden Markov modeling, then the structural recognizer analyzes them. The results from two recognizers are combined in a probabilistic framework complementing the Markov assumption of hidden Markov modeling. The experimental results showed significant performance improcements in error reduction and computation time as compared to the statistical approach alone.

Unconstrained keyword spotting using phone lattices with application to spoken document retrieval

Traditional hidden Markov model (HMM) word spotting requires both explicit HMM models of each desired keyword and a computationally expensive decoding pass. For certain applications, such as audio indexing or information retrieval, conventional word spotting may be too constrained or impractically slow. This paper presents an alternative technique, where a phone lattice—representing multiple phone hypotheses—is pre-computed prior to need. Given a phone decomposition of any desired keyword, the lattice may be rapidly searched to find putative occurrences of the keyword. Though somewhat less accurate, this can be substantially faster (orders of magnitude) and more flexible (any keyword may be detected) than previous approaches. This paper presents algorithms for lattice generation and scanning, and experimental results, including comparison with conventional keyword-HMM approaches. Finally, word spotting based on phone lattice scanning is demonstrated to be effective for spoken document retrieval.

Speech recognition using hidden Markov models with polynomial regression functions as nonstationary states

Proposes, implements, and evaluates a class of nonstationary-state hidden Markov models (HMMs) having each state associated with a distinct polynomial regression function of time plus white Gaussian noise. The model represents the transitional acoustic trajectories of speech in a parametric manner, and includes the standard stationary-state HMM as a special, degenerated case. The authors develop an efficient dynamic programming technique which includes the state sojourn time as an optimization variable, in conjunction with a state-dependent orthogonal polynomial regression method, for estimating the model parameters. Experiments on fitting models to speech data and on limited-vocabulary speech recognition demonstrate consistent superiority of these nonstationary-state HMMs over the traditional stationary-state HMMs.< <ETX xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">&gt;</ETX>