Download:
|
by D. J. Kershaw, A. J. Robinson, S. J. Renals
http://svr-www.eng.cam.ac.uk/~djk/Publications/kershaw.slt96.ps.gz
Add To MetaCart
Abstract:
Abbot is the hybrid connectionist-hidden Markov model large-vocabulary speech recognition system developed at Cambridge University. In this system, a recurrent network maps each acoustic vector to an estimate of the posterior probabilities of the phone classes. This paper describes the system which participated in the November 1995 ARPA H3 Multiple Unknown Microphones (MUM) evaluation of continuous speech recognition systems, under the guise of the CU-CON system. The emphasis of the paper is on the changes made to the 1994 Abbot system, specifically to accomodate the H3 task. This includes improved acoustic modelling using limited word-internal context-dependent models, training on the Wall Street Journal secondary channel database, the linear input network for speaker and environmental adaptation and the continued development of a realtime single-pass decoder well suited to the hybrid approach. Experimental results are reported for various test and development
Citations
|
317
|
Connectionist Speech Recognition: A Hybrid Approach
– Bourlard, Morgan
- 1994
|
|
305
|
Maximum Likelihood Linear Regression for speaker adaptation of continuous density Hidden Markov Models
– Leggetter, Woodland
- 1995
|
|
276
|
RASTA processing of speech
– Hermansky, Morgan
- 1994
|
|
147
|
An application of recurrent nets to phone probability estimation
– Robinson
- 1994
|
|
78
|
Boosting and other ensemble methods
– Drucker, Cortes, et al.
- 1994
|
|
44
|
benchmark tests for the arpa spoken language program
– Pallett, Fiscus, et al.
- 1993
|
|
43
|
An essential a ∗ stack decoder algorithm for continuous speech recognition with a stochastic language model
– Paul
- 1992
|
|
38
|
Context-dependent classes in a hybrid recurrent network-HMM speech recognition system
– Kershaw, Hochberg, et al.
- 1996
|
|
33
|
Efficient search using posterior phone probability estimators
– Renals, Hochberg
- 1995
|
|
29
|
Decoder technology for connectionist large vocabulary speech recognition
– Renals, Hochberg
- 1995
|
|
24
|
The Use of Recurrent Neural Networks in Continuous Speech Recognition
– ROBINSON, —RENALS
- 1995
|
|
23
|
Continuous speech recognition by connectionist statistical methods
– Bourlard, Morgan
- 1993
|
|
22
|
A tree search strategy for large vocabulary continuous speech recognition
– Gopalakrishnan, Bahl, et al.
- 1995
|
|
15
|
Connectionist model combination for large vocabulary speech recognition
– Hochberg, Cook, et al.
- 1994
|
|
15
|
Several Improvements to a Recurrent Error Propagation Network Phone Recognition System
– Robinson
- 1991
|
|
12
|
Unsupervised speaker-adaptation for hybrid HMM-MLP continuous speech recognition systems
– Neto, Almeida, et al.
- 1995
|
|
9
|
Learning phonological rule probabilities from speech corpora with exploratory computational phonology
– Tajchman, Jurafsky, et al.
- 1995
|
|
8
|
ABBOT: The CUED Hybrid ConnectionistHMM Large Vocabulary Recognition System
– Hochberg, Renals, et al.
- 1995
|
|
6
|
Backpropagation through time: what does it mean and how to do it
– Werbos
- 1990
|
|
2
|
Context-Dependent Modelling
– Kershaw, Hochberg, et al.
- 1995
|
|
1
|
Phone Deactivation Pruning
– Renals
- 1996
|
