Showing papers on "Hidden Markov model published in 1997"

Statistical methods for speech recognition

Abstract: The speech recognition problem hidden Markov models the acoustic model basic language modelling the Viterbi search hypothesis search on a tree and the fast match elements of information theory the complexity of tasks - the quality of language models the expectation - maximization algorithm and its consequences decision trees and tree language models phonetics from orthography - spelling-to-base from mappings triphones and allophones maximum entropy probability estimation and language models three applications of maximum entropy estimation to language modelling estimation of probabilities from counts and the Back-Off method.

Coupled hidden Markov models for complex action recognition

Abstract: We present algorithms for coupling and training hidden Markov models (HMMs) to model interacting processes, and demonstrate their superiority to conventional HMMs in a vision task classifying two-handed actions. HMMs are perhaps the most successful framework in perceptual computing for modeling and classifying dynamic behaviors, popular because they offer dynamic time warping, a training algorithm and a clear Bayesian semantics. However the Markovian framework makes strong restrictive assumptions about the system generating the signal-that it is a single process having a small number of states and an extremely limited state memory. The single-process model is often inappropriate for vision (and speech) applications, resulting in low ceilings on model performance. Coupled HMMs provide an efficient way to resolve many of these problems, and offer superior training speeds, model likelihoods, and robustness to initial conditions.

Minimum classification error rate methods for speech recognition

Abstract: A critical component in the pattern matching approach to speech recognition is the training algorithm, which aims at producing typical (reference) patterns or models for accurate pattern comparison. In this paper, we discuss the issue of speech recognizer training from a broad perspective with root in the classical Bayes decision theory. We differentiate the method of classifier design by way of distribution estimation and the discriminative method of minimizing classification error rate based on the fact that in many realistic applications, such as speech recognition, the real signal distribution form is rarely known precisely. We argue that traditional methods relying on distribution estimation are suboptimal when the assumed distribution form is not the true one, and that "optimality" in distribution estimation does not automatically translate into "optimality" in classifier design. We compare the two different methods in the context of hidden Markov modeling for speech recognition. We show the superiority of the minimum classification error (MCE) method over the distribution estimation method by providing the results of several key speech recognition experiments. In general, the MCE method provides a significant reduction of recognition error rate.

Learning and recognizing human dynamics in video sequences

Abstract: This paper describes a probabilistic decomposition of human dynamics at multiple abstractions, and shows how to propagate hypotheses across space, time, and abstraction levels. Recognition in this framework is the succession of very general low level grouping mechanisms to increased specific and learned model based grouping techniques at higher levels. Hard decision thresholds are delayed and resolved by higher level statistical models and temporal context. Low-level primitives are areas of coherent motion found by EM clustering, mid-level categories are simple movements represented by dynamical systems, and high-level complex gestures are represented by Hidden Markov Models as successive phases of ample movements. We show how such a representation can be learned from training data, and apply It to the example of human gait recognition.

Nymble: a High-Performance Learning Name-finder

Abstract: This paper presents a statistical, learned approach to finding names and other nonrecursive entities in text (as per the MUC-6 definition of the NE task), using a variant of the standard hidden Markov model. We present our justification for the problem and our approach, a detailed discussion of the model itself and finally the successful results of this new approach.

Monte Carlo Implementation of Gaussian Process Models for Bayesian Regression and Classification

Abstract: Gaussian processes are a natural way of defining prior distributions over functions of one or more input variables. In a simple nonparametric regression problem, where such a function gives the mean of a Gaussian distribution for an observed response, a Gaussian process model can easily be implemented using matrix computations that are feasible for datasets of up to about a thousand cases. Hyperparameters that define the covariance function of the Gaussian process can be sampled using Markov chain methods. Regression models where the noise has a t distribution and logistic or probit models for classification applications can be implemented by sampling as well for latent values underlying the observations. Software is now available that implements these methods using covariance functions with hierarchical parameterizations. Models defined in this way can discover high-level properties of the data, such as which inputs are relevant to predicting the response.

Two Methods for Improving Performance of a HMM and their Application for Gene Finding

Abstract: A hidden Markov model for gene finding consists of submodels for coding regions, splice sites, introns, intergenic regions and possibly more. It is described how to estimate the model as a whole from labeled sequences instead of estimating the individual parts independently from subsequences. It is argued that the standard maximum likelihood estimation criterion is not optimal for training such a model. Instead of maximizing the probability of the DNA sequence, one should maximize the probability of the correct prediction. Such a criterion, called conditional maximum likelihood, is used for the gene finder 'HMM-gene'. A new (approximative) algorithm is described, which finds the most probable prediction summed over all paths yielding the same prediction. We show that these methods contribute significantly to the high performance of HMMgene.

Probabilistic Independence Networks for Hidden Markov Probability Models

Abstract: Graphical techniques for modeling the dependencies of randomvariables have been explored in a variety of different areas includingstatistics, statistical physics, artificial intelligence, speech recognition, image processing, and genetics.Formalisms for manipulating these models have been developedrelatively independently in these research communities. In this paper weexplore hidden Markov models (HMMs) and related structures within the general framework of probabilistic independencenetworks (PINs). The paper contains a self-contained review of the basic principles of PINs.It is shown that the well-known forward-backward (F-B) and Viterbialgorithms for HMMs are special cases of more general inference algorithms forarbitrary PINs. Furthermore, the existence of inference and estimationalgorithms for more general graphical models provides a set of analysistools for HMM practitioners who wish to explore a richer class of HMMstructures.Examples of relatively complex models to handle sensorfusion and coarticulationin speech recognitionare introduced and treated within the graphical model framework toillustrate the advantages of the general approach.

Human action learning via hidden Markov model

Abstract: To successfully interact with and learn from humans in cooperative modes, robots need a mechanism for recognizing, characterizing, and emulating human skills. In particular, it is our interest to develop the mechanism for recognizing and emulating simple human actions, i.e., a simple activity in a manual operation where no sensory feedback is available. To this end, we have developed a method to model such actions using a hidden Markov model (HMM) representation. We proposed an approach to address two critical problems in action modeling: classifying human action-intent, and learning human skill, for which we elaborated on the method, procedure, and implementation issues in this paper. This work provides a framework for modeling and learning human actions from observations. The approach can be applied to intelligent recognition of manual actions and high-level programming of control input within a supervisory control paradigm, as well as automatic transfer of human skills to robotic systems.

LAFTER: lips and face real time tracker

Abstract: This paper describes an active-camera real-time system for tracking, shape description, and classification of the human face and mouth using only an SGI Indy computer. The system is based on use of 2-D blob features, which are spatially-compact clusters of pixels that are similar in terms of low-level image properties. Patterns of behavior (e.g., facial expressions and head movements) can be classified in real-time using Hidden Markov Model (HMM) methods. The system has been tested on hundreds of users and has demonstrated extremely reliable and accurate performance. Typical classification accuracies are near 100%.

Discriminative utterance verification for connected digits recognition

Abstract: Utterance verification represents an important technology in the design of user-friendly speech recognition systems It involves the recognition of keyword strings and the rejection of nonkeyword strings This paper describes a hidden Markov model-based (HMM-based) utterance verification system using the framework of statistical hypothesis testing The two major issues on how to design keyword and string scoring criteria are addressed For keyword verification, different alternative hypotheses are proposed based on the scores of antikeyword models and a general acoustic filler model For string verification, different measures are proposed with the objective of detecting nonvocabulary word strings and possibly erroneous strings (so-called putative errors) This paper also motivates the need for discriminative hypothesis testing in verification One such approach based on minimum classification error training is investigated in detail When the proposed verification technique was integrated into a state-of-the-art connected digit recognition system, the string error rate for valid digit strings was found to decrease by 57% when setting the rejection rate to 5% Furthermore, the system was able to correctly reject over 999% of nonvocabulary word strings

Finding Genes in DNA with a Hidden Markov Model

Abstract: This study describes a new Hidden Markov Model (HMM) system for segmenting uncharacterized genomic DNA sequences into exons, introns, and intergenic regions. Separate HMM modules were designed and trained for specific regions of DNA: exons, introns, intergenic regions, and splice sites. The models were then tied together to form a biologically feasible topology. The integrated HMM was trained further on a set of eukaryotic DNA sequences and tested by using it to segment a separate set of sequences. The resulting HMM system which is called VEIL (Viterbi Exon-Intron Locator), obtains an overall accuracy on test data of 92% of total bases correctly labelled, with a correlation coefficient of 0.73. Using the more stringent test of exact exon prediction, VEIL correctly located both ends of 53% of the coding exons, and 49% of the exons it predicts are exactly correct. These results compare favorably to the best previous results for gene structure prediction and demonstrate the benefits of using HMMs for this problem.

Isolated sign language recognition using hidden Markov models

Abstract: This paper is concerned with the video-based recognition of isolated signs. Concentrating on the manual parameters of sign language, the system aims for the signer dependent recognition of 262 different signs. For hidden Markov modelling a sign is considered a doubly stochastic process, represented by an unobservable state sequence. The observations emitted by the states are regarded as feature vectors, that are extracted from video frames. The system achieves recognition rates up to 94%.

meta-MEME: Motif-based hidden Markov models of protein families

Abstract: Motivation: Modeling families of related biological sequences using Hidden Markov models (HMMs), although increasingly widespread, faces at least one major problem: because of the complexity of these mathematical models, they require a relatively large training set in order to accurately recognize a given family. For families in which there are few known sequences, a standard linear HMM contains too many parameters to be trained adequately. Results: This work attempts to solve that problem by generating smaller HMMs which precisely model only the conserved regions of the family. These HMMs are constructed from motif models generated by the EM algorithm using the MEME software. Because motif-based HMMs have relatively few parameters, they can be trained using smaller data sets. Studies of short chain alcohol dehydrogenases and 4Fe-4S ferredoxins support the claim that motif-based HMMs exhibit increased sensitivity and selectivity in database searches, especially when training sets contain few sequences.

Learning topological maps with weak local odometric information

Abstract: Topological maps provide a useful abstraction for robotic navigation and planning. Although stochastic maps can theoretically be learned using the Baum-Welch algorithm, without strong prior constraint on the structure of the model it is slow to converge, requires a great deal of data, and is often stuck in local minima. In this paper, we consider a special case of hidden Markov models for robot-navigation environments, in which states are associated with points in a metric configuration space. We assume that the robot has some odometric ability to measure relative transformations between its configurations. Such odometry is typically not precise enough to suffice for building a global map, but it does give valuable local information about relations between adjacent states. We present an extension of the Baum-Welch algorithm that takes advantage of this local odometric information, yielding faster convergence to better solutions with less data.

Investigation of silicon auditory models and generalization of linear discriminant analysis for improved speech recognition

Abstract: Biologically motivated feature extraction algorithms have been found to provide significantly robust performance in speech recognition systems, in the presence of channel and noise degradation, when compared to the standard features such as mel-cepstrum coefficients. However, auditory feature extraction is computationally expensive, and makes these features useless for real-time speech recognition systems. In this thesis, I investigate the use of low power techniques and custom analog VLSI for auditory feature extraction. I first investigated the basilar-membrane model and the hair-cell model chips that were designed by Liu (Liu, 1992). I performed speech recognition experiments to evaluate how well these chips would perform as a front-end to a speech recognizer. Based on the experience gained by these experiments, I propose an alternate architecture that goes beyond the basilar-membrane model, and, using which, auditory features can be computed in real time. These chips have been designed and tested, and consume only a few milliwatts of power as compared to general purpose digital machines that consume several Watts. I have also investigated Linear Discriminant Analysis (LDA) for dimension reduction of auditory features. Researchers have used Fisher-Rao linear discriminant analysis (LDA) to reduce the feature dimension. They model the low-dimensional features obtained from LDA as the outputs of a Markov process with hidden states (HMM). I present a unified framework for the problem of dimension reduction and HMM parameter estimation by modeling the original features with reduced-rank HMM. This re-formulation also leads to a generalization of LDA that is consistent with the heteroscedastic state models used in HMM, and give better performance when tested on a digit recognition task.

Neural-network classifiers for recognizing totally unconstrained handwritten numerals

Abstract: Artificial neural networks have been recognized as a powerful tool for pattern classification problems, but a number of researchers have also suggested that straightforward neural-network approaches to pattern recognition are largely inadequate for difficult problems such as handwritten numeral recognition. In this paper, we present three sophisticated neural-network classifiers to solve complex pattern recognition problems: multiple multilayer perceptron (MLP) classifier, hidden Markov model (HMM)/MLP hybrid classifier, and structure-adaptive self-organizing map (SOM) classifier. In order to verify the superiority of the proposed classifiers, experiments were performed with the unconstrained handwritten numeral database of Concordia University, Montreal, Canada. The three methods have produced 97.35%, 96.55%, and 96.05% of the recognition rates, respectively, which are better than those of several previous methods reported in the literature on the same database.

Adapting hidden Markov models for ASL recognition by using three-dimensional computer vision methods

Abstract: We present an approach to continuous American sign language (ASL) recognition, which uses as input 3D data of arm motions. We use computer vision methods for 3D object shape and motion parameter extraction and an ascension technologies 'Flock of Birds' interchangeably to obtain accurate 3D movement parameters of ASL sentences, selected from a 53-sign vocabulary and a widely varied sentence structure. These parameters are used as features for hidden Markov models (HMMs). To address coarticulation effects and improve our recognition results, we experimented with two different approaches. The first consists of training context-dependent HMMs and is inspired by speech recognition systems. The second consists of modeling transient movements between signs and is inspired by the characteristics of ASL phonology. Our experiments verified that the second approach yields better recognition results.

The ORL active floor [sensor system]

Abstract: A novel type of sensor system called the active floor is presented that allows the time-varying spatial weight distribution of the active office environment to be captured. The properties of the active floor are described, showing that it differs substantially from other commonly encountered sensor systems. Furthermore, classification of the footstep signature of a number of individuals is attempted by application of the hidden Markov model technique.

Microphone array based speech recognition with different talker-array positions

Abstract: The use of a microphone array for hands-free continuous speech recognition in noisy and reverberant environment is investigated. An array of eight omnidirectional microphones was placed at different angles and distances from the talker. A time delay compensation module was used to provide a beamformed signal as input to a hidden Markov model (HMM) based recognizer. A phone HMM adaptation, based on a small amount of phonetically rich sentences, further improved the recognition rate obtained by applying only beamforming. These results were confirmed both by experiments conducted in a noisy and reverberant environment and by simulations. In the latter case, different conditions were recreated by using the image method to reproduce synthetic versions of the array microphone signals.

Decision-directed frame-synchronous adaptive equalization filtering of a speech signal by implementing a hidden markov model

Abstract: For equalizing a speech signal constituted by an observed sequence of successive input sound frames, which speech signal is liable to be affected by disturbances, the speech signal is modelled by means of a hidden Markov model and, at each instant t: equalization filters are constituted in association with the paths in the Markov sense at instant t; at least a plurality of the equalization filters are applied to the frames to obtain, at instant t, a plurality of filtered sound frame sequences and an utterance probability for each of the paths respectively associated with the equalization filters applied; the equalization filter corresponding to the most probable path in the Markov sense is selected; and the filtered frame supplied by the selected equalization filter is selected as the equalized frame.

Predicting protein structure using hidden markov models

Abstract: We discuss how methods based on hidden Markov models performed in the fold recognition section of the CASP2 experiment. Hidden Markov models were built for a set of about a thousand structures from the PDB database, and each CASP2 target sequence was scored against this library of hidden Markov models. In addition, a hidden Markov model was built for each of the target sequences, and all of the sequences in PDB were scored against that target model. Having high scores from both methods was found to be highly indicative of the target and a structure being homologous. Predictions were made based on several criteria: the scores with the structure models, the scores with the target models, consistency between the secondary structure in the known structure and predictions for the target (using the program PhD), human examination of predicted alignments between target and structure (using RASMOL), and solvation preferences in the alignment of the target and structure. The method worked well in comparison to other methods used at CASP2 for targets of moderate difficulty, where the closest structure in PDB could be aligned to the target with at least 15% residue identity. There was no evidence for the method''s effectiveness for harder cases, where the residue identity was much lower than 15%.

Learning dialogue strategies within the Markov decision process framework

Abstract: We introduce a stochastic model for dialogue systems based on the Markov decision process. Within this framework we show that the problem of dialogue strategy design can be stated as an optimization problem, and solved by a variety of methods, including the reinforcement learning approach. The advantages of this new paradigm include objective evaluation of dialogue systems and their automatic design and adaptation. We show some preliminary results on learning a dialogue strategy for an air travel information system.

Estimation of generalized multisensor hidden Markov chains and unsupervised image segmentation

Abstract: This paper attacks the problem of generalized multisensor mixture estimation. A distribution mixture is said to be generalized when the exact nature of components is not known, but each of them belongs to a finite known set of families of distributions. Estimating such a mixture entails a supplementary difficulty: one must label, for each class and each sensor, the exact nature of the corresponding distribution. Such generalized mixtures have been studied assuming that the components lie in the Pearson system. We propose a more general procedure with applications to estimating generalized multisensor hidden Markov chains. Our proposed method is applied to the problem of unsupervised image segmentation. The method proposed allows one to: 1) identify the conditional distribution for each class and each sensor, 2) estimate the unknown parameters in this distribution, 3) estimate priors, and 4) estimate the "true" class image.

Lipreading from color video

Abstract: We have designed and implemented a lipreading system that recognizes isolated words using only color video of human lips (without acoustic data). The system performs video recognition using "snakes" to extract visual features of geometric space, Karhunen-Loeve transform (KLT) to extract principal components in the color eigenspace, and hidden Markov models (HMM's) to recognize the combined visual features sequences. With the visual information alone, we were able to achieve 94% accuracy for ten isolated words.

Estimation of generalized mixtures and its application in image segmentation

Abstract: We introduce the notion of a generalized mixture and propose some methods for estimating it, along with applications to unsupervised statistical image segmentation. A distribution mixture is said to be "generalized" when the exact nature of the components is not known, but each belongs to a finite known set of families of distributions. For instance, we can consider a mixture of three distributions, each being exponential or Gaussian. The problem of estimating such a mixture contains thus a new difficulty: we have to label each of three components (there are eight possibilities). We show that the classical mixture estimation algorithms-expectation-maximization (EM), stochastic EM (SEM), and iterative conditional estimation (ICE)-can be adapted to such situations once as we dispose of a method of recognition of each component separately. That is, when we know that a sample proceeds from one family of the set considered, we have a decision rule for what family it belongs to. Considering the Pearson system, which is a set of eight families, the decision rule above is defined by the use of "skewness" and "kurtosis". The different algorithms so obtained are then applied to the problem of unsupervised Bayesian image segmentation, We propose the adaptive versions of SEM, EM, and ICE in the case of "blind", i.e., "pixel by pixel", segmentation. "Global" segmentation methods require modeling by hidden random Markov fields, and we propose adaptations of two traditional parameter estimation algorithms: Gibbsian EM (GEM) and ICE allowing the estimation of generalized mixtures corresponding to Pearson's system. The efficiency of different methods is compared via numerical studies, and the results of unsupervised segmentation of three real radar images by different methods are presented.