Welcome to Scribd, the world's digital library. Read, publish, and share books and documents. See more
Download
Standard view
Full view
of .
Save to My Library
Look up keyword or section
Like this
2Activity

Table Of Contents

0 of .
Results for:
No results containing your search query
P. 1
(Brain Study)_Speech Recognition Using Neural Networks

(Brain Study)_Speech Recognition Using Neural Networks

Ratings: (0)|Views: 7 |Likes:
Published by nick Hall

More info:

Published by: nick Hall on Oct 03, 2011
Copyright:Attribution Non-commercial

Availability:

Read on Scribd mobile: iPhone, iPad and Android.
download as PDF, TXT or read online from Scribd
See more
See less

11/05/2013

pdf

text

original

 
Speech Recognition using Neural Networks
Joe Tebelskis
May 1995CMU-CS-95-142School of Computer ScienceCarnegie Mellon UniversityPittsburgh, Pennsylvania 15213-3890
Submitted in partial fulfillment of the requirements for a degree of Doctor of Philosophy in Computer Science
Thesis Committee:
Alex Waibel, chairRaj ReddyJaime CarbonellRichard Lippmann, MIT Lincoln LabsCopyright
©1995
Joe Tebelskis
This research was supported during separate phases by ATR Interpreting Telephony Research Laboratories,NEC Corporation, Siemens AG, the National Science Foundation, the Advanced Research Projects Adminis-tration, and the Department of Defense under Contract No. MDA904-92-C-5161.The views and conclusions contained in this document are those of the author and should not be interpreted asrepresenting the official policies, either expressed or implied, of ATR, NEC, Siemens, NSF, or the UnitedStates Government.
 
Keywords
: Speech recognition, neural networks, hidden Markov models, hybrid systems,acoustic modeling, prediction, classification, probability estimation, discrimination, globaloptimization.
 
iii
Abstract
This thesis examines how artificial neural networks can benefit a large vocabulary, speakerindependent, continuous speech recognition system. Currently, most speech recognitionsystems are based on hidden Markov models (HMMs), a statistical framework that supportsboth acoustic and temporal modeling. Despite their state-of-the-art performance, HMMsmake a number of suboptimal modeling assumptions that limit their potential effectiveness.Neural networks avoid many of these assumptions, while they can also learn complex func-tions, generalize effectively, tolerate noise, and support parallelism. While neural networkscan readily be applied to acoustic modeling, it is not yet clear how they can be used for tem-poral modeling. Therefore, we explore a class of systems called
NN-HMM hybrids
, in whichneural networks perform acoustic modeling, and HMMs perform temporal modeling. Weargue that a NN-HMM hybrid has several theoretical advantages over a pure HMM system,including better acoustic modeling accuracy, better context sensitivity, more natural dis-crimination, and a more economical use of parameters. These advantages are confirmedexperimentally by a NN-HMM hybrid that we developed, based on context-independentphoneme models, that achieved 90.5% word accuracy on the Resource Management data-base, in contrast to only 86.0% accuracy achieved by a pure HMM under similar conditions.In the course of developing this system, we explored two different ways to use neural net-works for acoustic modeling: prediction and classification. We found that predictive net-works yield poor results because of a lack of discrimination, but classification networksgave excellent results. We verified that, in accordance with theory, the output activations of a classification network form highly accurate estimates of the posterior probabilities
P
(
class
|
input 
), and we showed how these can easily be converted to likelihoods
P
(
input 
|
class
) for standard HMM recognition algorithms. Finally, this thesis reports how weoptimized the accuracy of our system with many natural techniques, such as expanding theinput window size, normalizing the inputs, increasing the number of hidden units, convert-ing the network’s output activations to log likelihoods, optimizing the learning rate scheduleby automatic search, backpropagating error from word level outputs, and using genderdependent networks.

Activity (2)

You've already reviewed this. Edit your review.
1 thousand reads
1 hundred reads

You're Reading a Free Preview

Download
/*********** DO NOT ALTER ANYTHING BELOW THIS LINE ! ************/ var s_code=s.t();if(s_code)document.write(s_code)//-->