SRI's DECIPHER system

Murveit, Hy; Cohen, Michael A.; Price, Patti; Weintraub, M.; Bernstein, Jared

doi:10.3115/100964.100990

Cited by 22 publications

(6 citation statements)

References 9 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…The results listed in Table I are approximately the same as those achieved by more conventional systems tested on the same data [13,14,15,16] and the perplexity 60 grammar. Given the difficulty of the task and the early stage of development of this system, however, we consider these results quite respectable.…”

Section: Interpretation Of the Resultssupporting

confidence: 60%

See 1 more Smart Citation

Continuous speech recognition from phonetic transcription

Levinson

Ljolje

1989

Proceedings of the Workshop on Speech and Natural Language - HLT '89

View full text Add to dashboard Cite

Previous research by the authors has been directed toward phonetic transcription of fluent speech. We have applied our techniques to speech recognition on the DARPA Resource Management Task. In order to perform speech recognition, however, the phonetic transcription must be interpreted as a sequence of words. A central component of this process is lexical access for which a novel method is proposed.

show abstract

Section: Interpretation Of the Resultssupporting

confidence: 60%

“…The word accuracy of our system is not as good as that obtained on exactly the same data by several other conventional systems [13,14,15,16]. However, we believe that a few correctable shortcomings of the existing system are responsible for the disparity.…”

Section: Introductioncontrasting

confidence: 68%

Continuous speech recognition from phonetic transcription

Levinson

Ljolje

1989

Proceedings of the Workshop on Speech and Natural Language - HLT '89

View full text Add to dashboard Cite

show abstract

“…[4], because this design can satisfy the requirements of trainable and stability at the same time. A typical speech recognition front end [5] is shown in Figure 3. Here, the speech signal is filtered, windowed, and then transform coded, typically with mel-frequency weighted cepstrals.…”

Section: Speech Recognition Component In Pervasive Hmi Layermentioning

confidence: 99%

A Pervasive Computing System Design and Implementation for Mandarin e-Learning

Zhang

2006

2006 First International Symposium on Pervasive Computing and Applications

View full text Add to dashboard Cite

Pervasive computing has a great potential for many next-generation IT applications. This paper describes a pervasive computing system design and implementation for mandarin e-learning. We propose a Human-centered Pervasive Computing System Model (HPC), a Layered Architecture Analysis and Design Method (LAAD). Based on the HPC model and the LAAD method, a pervasive computing based mandarin e-learning system is designed and implemented. The design and implementation issues are discussed in more details.

show abstract

“…A number of sites have already moved in this direction by changing from contextindependent phone (monophone) models to left and right context dependent phone (triphone) models [18]. h further step along this line has been the inclusion of crossword triphone models [7,10,13] which has minimized the ability of the acoustic models to learn the bigram language model. 1 These changes have improved recognition performance when trained and tested on the same database, but their effects on vocabulary independence have not been tested.…”

Section: Els [5]mentioning

confidence: 99%

On the interaction between true source, training, and testing language models

Paul¹,

Baker²

1990

Proceedings of the Workshop on Speech and Natural Language - HLT '90

View full text Add to dashboard Cite

An interaction has been found between the true source language model, training language model, and the testing language model. This interaction has implications for vocabulary independent modeling, testing methodologies, discriminative training, and the adequacy of our current databases for continuous speech recognition (CSR) development. The current DARPA databases suffer from the described difficulties which suggests that new CSR databases are needed if we are to further advance the state-of-the-art. The Interaction During Training When a category model (e.g. a context-free (CF) model such as a monophone) is used to a model a set of subcategories (e.g. context-dependent (CD) models such as triphones), the category model becomes the subcategory prior-probability weighted average of the subcategory models: Meat E PsubeatMsubcat where M denotes a model. (The mathematics used here are intended to be conceptual rather than rigorous. Thus models will be considered to be averages. In practice, the method for deriving a model from a set of sub-models or observations is highly dependent upon the form of model used.) In a field, such as speech recognition, where models are trained from exemplars, the subcategory model will generally be: N 1 Msttbcat = ~ ~.= Osubeat,i where 08=bcat,i is an observation emitted from the subcategory. Mcat combines both the subcategory models and the prior-probability of the subcategories and similarly Msubcat combines the observations and their (sampled) prior-probabilities.

show abstract

SRI's DECIPHER system

Cited by 22 publications

References 9 publications

Continuous speech recognition from phonetic transcription

Continuous speech recognition from phonetic transcription

A Pervasive Computing System Design and Implementation for Mandarin e-Learning

On the interaction between true source, training, and testing language models

Contact Info

Product

Resources

About