Corpus ID: 58576075

The Use of Context in Large Vocabulary Speech Recognition

@inproceedings{Odell1995TheUO,
  title={The Use of Context in Large Vocabulary Speech Recognition},
  author={J. J. Odell},
  year={1995}
}
In recent years, considerable progress has been made in the eld of continuous speech recognition where the predominant technology is based on hidden Markov models (HMMs). HMMs represent sequences of time varying speech spectra using probabilistic functions of an underlying Markov chain. However, because the probability distribution represented by a HMM is very simple, its discriminative ability is limited. As a consequence, a careful choice of the units represented by each model is required in… Expand
Reformulating HMM as a trajectory model by imposing explicit relationships between static and dynamic features
TLDR
A novel acoustic model, named a trajectory HMM, is described, derived from the HMM whose state output vector includes both static and dynamic features, which can improve the performance of continuous speech recognition and speech synthesis experiments. Expand
Hidden Model Sequence Models for Automatic Speech Recognition
TLDR
The novel framework of hidden model sequence modelling using the basis of hidden Markov models is presented and special attention is given to the modelling of sub-phone deletions. Expand
Fast and accurate acoustic modelling with semi-continuous HMMs
TLDR
A node splitting criterion appropriate for SC-HMMs is introduced: it is based on a distance measure between the mixtures of gaussian pdfs as involved in SC-HMM state modelling, which contrasts with other criteria from literature which are based on simplified pdfs to manage the algorithmic complexity. Expand
Natural statistical models for automatic speech recognition
TLDR
A new method for automatic speech recognition is developed where the natural statistical properties of speech are used to determine the probabilistic model, and can be seen as a general discriminative structure-learning procedure for Bayesian networks. Expand
Towards knowledge-based features for HMM based large vocabulary automatic speech recognition
TLDR
This paper has engineered an approach where neural networks are trained to map short-term spectral features to the posterior probability of some distinctive features of the speech signal, and uses these probabilities as features in a large vocabulary tied-state HMM-based recognizer. Expand
Example based continuous speech recognition
TLDR
A complete example based system for continuous speech recognition is developed, focusing on example based acoustic modelling, on a new model for the prior probability of recognition hypotheses and on the development of an efficient search algorithm. Expand
An efficient search space representation for large vocabulary continuous speech recognition
TLDR
A memory-efficient search topology that enables the use of such detailed acoustic and language models in a one pass time-synchronous recognition system and allows great flexibility in the choice of LMs. Expand
Speech recognition with speech synthesis models by marginalising over decision tree leaves
TLDR
A simple approach that generates ASR models from a trained set of TTS models by marginalising over the TTS contexts that are not used by ASR is proposed. Expand
Speech recognition based on statistical models including multiple phonetic decision trees
TLDR
An optimization technique of state sequences based on an annealing process using multiple decision trees is proposed, and the deterministicAnnealing expectation maximization algorithm is used as the training algorithm. Expand
Data selection and model combination in connectionist speech recognition
TLDR
Results on the optical character recognition task suggest that boosting can provide considerable gains in classification performance, and the application of boosting to acoustic modelling has been investigated, and a modified boosting procedure developed. Expand
...
1
2
3
4
5
...

References

SHOWING 1-10 OF 125 REFERENCES
The development of the 1994 HTK large vocabulary speech recognition system
TLDR
The HTK large vocabulary continuous speech recognition system uses tied-state crossword context-dependent mixture Gaussian HMMs and a dynamic network decoder that can operate in a single pass to allow exible and eecient system development, as well as multi-pass operation for use with computationally expensive acoustic and/or language models. Expand
Continuously variable duration hidden Markov models for automatic speech recognition
TLDR
The solution proposed here is to replace the probability distributions of duration with continuous probability density functions to form a continuously variable duration hidden Markov model (CVDHMM) which is ideally suited to specification of the durational density. Expand
Complexity reduction in a large vocabulary speech recognizer
The authors provide a detailed description of all aspects of the implementation of a large-vocabulary speaker-independent, continuous speech recognizer used as a tool for the development ofExpand
A One Pass Decoder Design For Large Vocabulary Recognition
TLDR
This paper shows that time-synchronous one-pass decoding using cross-word triphones and a trigram language model can be implemented using a dynamically built tree-structured network and is relatively efficient in implementation. Expand
Tied mixture continuous parameter models for large vocabulary isolated speech recognition
TLDR
A class of very general hidden Markov models which can accommodate sequences of information-bearing acoustic feature vectors lying either in a discrete or in a continuous space are considered. Expand
The acoustic-modeling problem in automatic speech recognition
TLDR
This thesis is primarily concerned with the use of hidden Markov models to model sequences of feature vectors which lie in a continuous space such as R sub N and explores the trade-off between packing a lot of information into such sequences and being able to model them accurately. Expand
Using MAP estimated parameters to improve HMM speech recognition performance
TLDR
A maximum a posteriori (MAP) parameter estimation approach for improving the state-duration modeling capability and incorporating a priori knowledge about the word-duration distribution into an HMM system is presented. Expand
A fast match for continuous speech recognition using allophonic models
TLDR
A highly accurate fast acoustic match for continuous speech recognition using allophonic models and efficient search techniques to select a set of candidate words from the vocabulary that match well with a given acoustic input. Expand
The 1994 Abbot hybrid connectionist-HMM large vocabulary recognition system.
TLDR
The emphasis of the paper is on the differences between the 1993 and 1994 versions of the ABBOT system, which includes the utilization of a larger training corpus, the extension of the lexicon, the application of a trigram language model, and the development of a near-realtime single-pass decoder well suited for the hybrid approach. Expand
Implementation Aspects of Large Vocabulary Recognition Based on Intraword and Interword Phonetic Units
TLDR
An architecture is considered in which several well known procedures are incorporated with some new ideas to implement a search procedure which maintains the accuracy of the full search but which can decode a single sentence in about one minute of computing time on a vectorized, concurrent processor. Expand
...
1
2
3
4
5
...