Neural Networks or Hidden Markov Models for Automatic Speech Recognition: Is there a Choice?

1992 ◽  
pp. 225-236 ◽  
Author(s):  
John S. Bridle
Keyword(s):  
Neural Networks ◽  
Speech Recognition ◽  
Hidden Markov Models ◽  
Automatic Speech Recognition ◽  
Markov Models ◽  
Hidden Markov

Literacy by Way of Automatic Speech Recognition

2011 ◽  
pp. 2074-2118
Author(s):  
Russell Gluck ◽  
John Fulcher
Keyword(s):  
Neural Networks ◽  
Speech Recognition ◽  
Automatic Speech Recognition ◽  
Markov Models ◽  
Hidden Markov ◽  
Time Warping ◽  
Oral Storytelling ◽  
Pattern Recognition Techniques ◽  
Dynamic Time ◽  
Over Time

The chapter commences with an overview of automatic speech recognition (ASR), which covers not only the de facto standard approach of hidden Markov models (HMMs), but also the tried-and-proven techniques of dynamic time warping and artificial neural networks (ANNs). The coverage then switches to Gluck’s (2004) draw-talk-write (DTW) process, developed over the past two decades to assist non-text literate people become gradually literate over time through telling and/or drawing their own stories. DTW has proved especially effective with “illiterate” people from strong oral, storytelling traditions. The chapter concludes by relating attempts to date in automating the DTW process using ANN-based pattern recognition techniques on an Apple Macintosh G4™ platform.

A CONNECTIONIST APPROACH TO SPEECH RECOGNITION

1993 ◽  
Vol 07 (04) ◽  
pp. 647-667 ◽  
Author(s):  
YOSHUA BENGIO
Keyword(s):  
Neural Networks ◽  
Artificial Neural Networks ◽  
Speech Recognition ◽  
Hidden Markov Models ◽  
Markov Models ◽  
Hidden Markov ◽  
Statistical Techniques ◽  
Recurrent Networks ◽  
Phoneme Recognition ◽  
Connectionist Approach

The task discussed in this paper is that of learning to map input sequences to output sequences. In particular, problems of phoneme recognition in continuous speech are considered, but most of the discussed techniques could be applied to other tasks, such as the recognition of sequences of handwritten characters. The systems considered in this paper are based on connectionist models, or artificial neural networks, sometimes combined with statistical techniques for recognition of sequences of patterns, stressing the integration of prior knowledge and learning. Different architectures for sequence and speech recognition are reviewed, including recurrent networks as well as hybrid systems involving hidden Markov models.

Sign in / Sign up

Export Citation Format

Share Document