LECTURE 26: PRACTICAL ISSUES · 2002. 3. 18. · LECTURE 26: PRACTICAL ISSUES Objectives: Discuss...

Return to Main

Objectives

Topolgy: Transition Matrix DTW Analogy Alternatives

Duration Modeling: State Duration Probabilities Number of States

Training Schedules: Simple Schedule More Complex Schedules Complex Models

On-Line Resources: Training Workshop Using HMMs The HTK Book

LECTURE 26: PRACTICALISSUES

Objectives:

Discuss common modeltopologies

❍

Provide model designguidelines

❍

Introduce typical trainingschedules

❍

●

This lecture combines material fromthis paper:

J. Picone, "Continuous SpeechRecognition Using HiddenMarkov Models", IEEE ASSPMagazine, vol. 7, no. 3, pp.26-41, July 1990.

LECTURE 26: PRACTICAL ISSUES

http://www.isip.msstate.edu/publications/courses/ece_8463/lectures/current/lecture_26/index.html (1 of 2) [3/17/2002 9:53:45 PM]

http://www.cs.ndsu.nodak.edu/~kvanhorn/HTKv22Book.ps

and information found in thisworkshop:

Speech Recognition SystemTraining Workshop, Institute forSignal and InformationProcessing, Mississippi StateUniversity, Mississippi State,Mississippi, 39762, USA, January2002.


http://www.isip.msstate.edu/publications/courses/ece_8463/lectures/current/lecture_26/index.html (2 of 2) [3/17/2002 9:53:45 PM]

http://www.isip.msstate.edu/conferences/current/program/session_06/http://www.isip.msstate.edu/conferences/current/program/session_06/

Return to Main

Introduction:

01: Organization (html, pdf)

Speech Signals:

02: Production (html, pdf)

03: Digital Models (html, pdf)

04: Perception (html, pdf)

05: Masking (html, pdf)

06: Phonetics and Phonology (html, pdf)

07: Syntax and Semantics (html, pdf)

Signal Processing:

08: Sampling (html, pdf)

09: Resampling (html, pdf)

10: Acoustic Transducers (html, pdf)

11: Temporal Analysis (html, pdf)

12: Frequency Domain Analysis (html, pdf)

13: Cepstral Analysis (html, pdf)

14: Exam No. 1 (html, pdf)

15: Linear Prediction (html, pdf)

16: LP-Based Representations

ECE 8463: FUNDAMENTALS OFSPEECH RECOGNITION

Professor Joseph PiconeDepartment of Electrical and Computer Engineering

Mississippi State University

email: [email protected]/fax: 601-325-3149; office: 413 Simrall

URL: http://www.isip.msstate.edu/resources/courses/ece_8463

Modern speech understanding systems merge interdisciplinary technologies from SignalProcessing, Pattern Recognition, Natural Language, and Linguistics into a unifiedstatistical framework. These systems, which have applications in a wide range of signalprocessing problems, represent a revolution in Digital Signal Processing (DSP). Once afield dominated by vector-oriented processors and linear algebra-based mathematics, thecurrent generation of DSP-based systems rely on sophisticated statistical modelsimplemented using a complex software paradigm. Such systems are now capable ofunderstanding continuous speech input for vocabularies of hundreds of thousands ofwords in operational environments.

In this course, we will explore the core components of modern statistically-based speechrecognition systems. We will view speech recognition problem in terms of three tasks:signal modeling, network searching, and language understanding. We will conclude ourdiscussion with an overview of state-of-the-art systems, and a review of availableresources to support further research and technology development.

Tar files containing a compilation of all the notes are available. However, these files arelarge and will require a substantial amount of time to download. A tar file of the htmlversion of the notes is available here. These were generated using wget:

wget -np -k -mhttp://www.isip.msstate.edu/publications/courses/ece_8463/lectures/current

A pdf file containing the entire set of lecture notes is available here. These were generatedusing Adobe Acrobat.

Questions or comments about the material presented here can be directed [email protected].

ECE 8463: FUNDAMENTALS OF SPEECH RECOGNITION

http://www.isip.msstate.edu/publications/courses/ece_8463/lectures/current/ (1 of 2) [3/17/2002 9:53:46 PM]

http://www.isip.msstate.edu/publications/courses/ece_8463/http://www.isip.msstate.edu/publications/courses/ece_8463/lectures/current/lecture_01/index.htmlhttp://www.isip.msstate.edu/publications/courses/ece_8463/lectures/current/lecture_01/lecture_01.pdfhttp://www.isip.msstate.edu/publications/courses/ece_8463/lectures/current/lecture_02/index.htmlhttp://www.isip.msstate.edu/publications/courses/ece_8463/lectures/current/lecture_02/lecture_02.pdfhttp://www.isip.msstate.edu/publications/courses/ece_8463/lectures/current/lecture_03/index.htmlhttp://www.isip.msstate.edu/publications/courses/ece_8463/lectures/current/lecture_03/lecture_03.pdfhttp://www.isip.msstate.edu/publications/courses/ece_8463/lectures/current/lecture_04/index.htmlhttp://www.isip.msstate.edu/publications/courses/ece_8463/lectures/current/lecture_04/lecture_04.pdfhttp://www.isip.msstate.edu/publications/courses/ece_8463/lectures/current/lecture_05/index.htmlhttp://www.isip.msstate.edu/publications/courses/ece_8463/lectures/current/lecture_05/lecture_05.pdfhttp://www.isip.msstate.edu/publications/courses/ece_8463/lectures/current/lecture_06/index.htmlhttp://www.isip.msstate.edu/publications/courses/ece_8463/lectures/current/lecture_06/lecture_06.pdfhttp://www.isip.msstate.edu/publications/courses/ece_8463/lectures/current/lecture_07/index.htmlhttp://www.isip.msstate.edu/publications/courses/ece_8463/lectures/current/lecture_07/lecture_07.pdfhttp://www.isip.msstate.edu/publications/courses/ece_8463/lectures/current/lecture_08/index.htmlhttp://www.isip.msstate.edu/publications/courses/ece_8463/lectures/current/lecture_08/lecture_08.pdfhttp://www.isip.msstate.edu/publications/courses/ece_8463/lectures/current/lecture_09/index.htmlhttp://www.isip.msstate.edu/publications/courses/ece_8463/lectures/current/lecture_09/lecture_09.pdfhttp://www.isip.msstate.edu/publications/courses/ece_8463/lectures/current/lecture_10/index.htmlhttp://www.isip.msstate.edu/publications/courses/ece_8463/lectures/current/lecture_10/lecture_10.pdfhttp://www.isip.msstate.edu/publications/courses/ece_8463/lectures/current/lecture_11/index.htmlhttp://www.isip.msstate.edu/publications/courses/ece_8463/lectures/current/lecture_11/lecture_11.pdfhttp://www.isip.msstate.edu/publications/courses/ece_8463/lectures/current/lecture_12/index.htmlhttp://www.isip.msstate.edu/publications/courses/ece_8463/lectures/current/lecture_12/lecture_12.pdfhttp://www.isip.msstate.edu/publications/courses/ece_8463/lectures/current/lecture_13/index.htmlhttp://www.isip.msstate.edu/publications/courses/ece_8463/lectures/current/lecture_13/lecture_13.pdfhttp://www.isip.msstate.edu/publications/courses/ece_8463/lectures/current/lecture_14/index.htmlhttp://www.isip.msstate.edu/publications/courses/ece_8463/lectures/current/lecture_14/lecture_14.pdfhttp://www.isip.msstate.edu/publications/courses/ece_8463/lectures/current/lecture_15/index.htmlhttp://www.isip.msstate.edu/publications/courses/ece_8463/lectures/current/lecture_15/lecture_15.pdfhttp://www.isip.msstate.edu/resources/courses/ece_8463http://www.isip.msstate.edu/publications/courses/ece_8463/lectures/current/lecture_all/notes.html.tar.gzhttp://www.isip.msstate.edu/publications/courses/ece_8463/lectures/current/lecture_all/notes.pdf.tar.gzmailto:[email protected]

(html, pdf)

17: Spectral Normalization (html, pdf)

Parameterization:

18: Differentiation (html, pdf)

19: Principal Components (html, pdf)

20: Linear Discriminant Analysis (html, pdf)

Acoustic Modeling:

21: Dynamic Programming (html, pdf)

22: Markov Models (html, pdf)

23: Parameter Estimation (html, pdf)

24: HMM Training (html, pdf)

25: Continuous Mixtures (html, pdf)

26: Practical Issues (html, pdf)

27: Decision Trees (html, pdf)

28: Limitations of HMMs (html, pdf)

Language Modeling:

ECE 8463: FUNDAMENTALS OF SPEECH RECOGNITION

http://www.isip.msstate.edu/publications/courses/ece_8463/lectures/current/ (2 of 2) [3/17/2002 9:53:46 PM]

http://www.isip.msstate.edu/publications/courses/ece_8463/lectures/current/lecture_16/index.htmlhttp://www.isip.msstate.edu/publications/courses/ece_8463/lectures/current/lecture_16/lecture_16.pdfhttp://www.isip.msstate.edu/publications/courses/ece_8463/lectures/current/lecture_17/index.htmlhttp://www.isip.msstate.edu/publications/courses/ece_8463/lectures/current/lecture_17/lecture_17.pdfhttp://www.isip.msstate.edu/publications/courses/ece_8463/lectures/current/lecture_18/index.htmlhttp://www.isip.msstate.edu/publications/courses/ece_8463/lectures/current/lecture_18/lecture_18.pdfhttp://www.isip.msstate.edu/publications/courses/ece_8463/lectures/current/lecture_19/index.htmlhttp://www.isip.msstate.edu/publications/courses/ece_8463/lectures/current/lecture_19/lecture_19.pdfhttp://www.isip.msstate.edu/publications/courses/ece_8463/lectures/current/lecture_20/index.htmlhttp://www.isip.msstate.edu/publications/courses/ece_8463/lectures/current/lecture_20/lecture_20.pdfhttp://www.isip.msstate.edu/publications/courses/ece_8463/lectures/current/lecture_21/index.htmlhttp://www.isip.msstate.edu/publications/courses/ece_8463/lectures/current/lecture_21/lecture_21.pdfhttp://www.isip.msstate.edu/publications/courses/ece_8463/lectures/current/lecture_22/index.htmlhttp://www.isip.msstate.edu/publications/courses/ece_8463/lectures/current/lecture_22/lecture_22.pdfhttp://www.isip.msstate.edu/publications/courses/ece_8463/lectures/current/lecture_23/index.htmlhttp://www.isip.msstate.edu/publications/courses/ece_8463/lectures/current/lecture_23/lecture_23.pdfhttp://www.isip.msstate.edu/publications/courses/ece_8463/lectures/current/lecture_24/index.htmlhttp://www.isip.msstate.edu/publications/courses/ece_8463/lectures/current/lecture_24/lecture_24.pdfhttp://www.isip.msstate.edu/publications/courses/ece_8463/lectures/current/lecture_25/index.htmlhttp://www.isip.msstate.edu/publications/courses/ece_8463/lectures/current/lecture_25/lecture_25.pdfhttp://www.isip.msstate.edu/publications/courses/ece_8463/lectures/current/lecture_26/lecture_26.pdfhttp://www.isip.msstate.edu/publications/courses/ece_8463/lectures/current/lecture_27/index.htmlhttp://www.isip.msstate.edu/publications/courses/ece_8463/lectures/current/lecture_27/lecture_27.pdfhttp://www.isip.msstate.edu/publications/courses/ece_8463/lectures/current/lecture_28/index.htmlhttp://www.isip.msstate.edu/publications/courses/ece_8463/lectures/current/lecture_28/lecture_28.pdf


Objectives:

Discuss common model topologies❍

Provide model design guidelines❍

Introduce typical training schedules❍

●

This lecture combines material from this paper:

J. Picone, "Continuous Speech RecognitionUsing Hidden Markov Models", IEEE ASSPMagazine, vol. 7, no. 3, pp. 26-41, July 1990.

and information found in this workshop:


http://www.isip.msstate.edu/publications/courses/ece_8463/lectures/current/lecture_26/lecture_26_00.html (1 of 2) [3/17/2002 9:53:46 PM]

Speech Recognition System TrainingWorkshop, Institute for Signal andInformation Processing, Mississippi StateUniversity, Mississippi State, Mississippi,39762, USA, January 2002.



http://www.isip.msstate.edu/conferences/current/program/session_06/http://www.isip.msstate.edu/conferences/current/program/session_06/

MODEL TOPOLOGYSRSTW'00: SESSION 06 - ACOUSTIC MODELING

http://www.isip.msstate.edu/conferences/srstw/program/session_06/acoustic_modeling/html/am_08.html [3/17/2002 9:53:47 PM]

THE DTW ANALOGYLECTURE 26: PRACTICAL ISSUES


Note the similarity to DTW with slopeconstraints

●



ALTERNATIVE MODEL TOPOLOGIESLECTURE 26: PRACTICAL ISSUES


Note the similarity to DTW with slopeconstraints

●



STATE DURATION PROBABILITIESLECTURE 26: PRACTICAL ISSUES

http://www.isip.msstate.edu/publications/courses/ece_8463/lectures/current/lecture_26/lecture_26_03.html [3/17/2002 9:53:48 PM]

DURATION CONSIDERATIONS ANDCLUSTERING

For word model-based systems, we often willconsider the duration of the word whenassigning the initial number of states in themodel:

●



Clustering approaches can be used to learnpronunciation variants:

●



TYPICAL TRAINING SCHEDULESLECTURE 26: PRACTICAL ISSUES


MORE EXTENSIVE TRAININGSCHEDULES

Phone-based HMM systems require a moreextensive training process.

●

Mixture generation is usually done last, and isperformed using a cluster-splitting approach.

●

Retraining after mixture generation isimportant.

●



MODEL COMPLEXITYLECTURE 26: PRACTICAL ISSUES


Return to Main

Sunday: 1.1: Registration 1.2: Reception

Monday: 2.1: Introduction 3.1: Foundation Classes 3.2: Programming 3.3: Algorithms

Tuesday: 4.1: Signal Processing 5.1: Transformations 5.2: Evaluations 5.3: Digits

Wednesday: 6.1: Acoustic Modeling 7.1: Model Design 7.2: Training 7.3: Alphadigits

Thursday: 8.1: Language Modeling 9.1: Hierarchical Search 9.2: N-gram Models 9.3: Rescoring

Friday: 10.1: LVCSR Systems 11.1: Conversational Systems 11.2: Building Systems 11.3: Switchboard

Resources: 12.1: Internet 12.2: Participants

WORKSHOP PROGRAM

This workshop is the first in a series of training workshops intended to train entry-levelresearchers on the details of building speech recognition systems using the ISIP system. Becauseseating is limited, preference will be given to graduate students planning on using the ISIPsystem in their research. ISIP will subsidize the travel expenses for students planning to attendthe workshop. Please see the registration page for more details.

We are now in the second year of our project to build a public domain system that is competitivewith state of the art. An overview of our mission to develop Internet Accessible SpeechRecognition Technology is available on the web. There is also a web site devoted todissemination of information, and a mailing list used to promote discussions within our usercommunity.

A preliminary agenda for the workshop is shown below. If you have comments or suggestionsabout the agenda, please feel free to contact us. We look forward to seeing you at SRSTW'00.

Location:

MorningLectures:

SimrallAuditorium,SimrallEngineringBuilding

AfternoonLaboratories:

ELI/ GilesRooms,MitchellMemorialLibrary

SRSTW'00: TECHNICAL PROGRAM

http://www.isip.msstate.edu/conferences/srstw/program/index.html (1 of 4) [3/17/2002 9:53:50 PM]

http://www.isip.msstate.edu/conferences/srstw/program/session_01/registration/index.htmlhttp://www.isip.msstate.edu/conferences/srstw/program/session_01/reception/index.htmlhttp://www.isip.msstate.edu/conferences/srstw/program/session_02/introduction/index.htmlhttp://www.isip.msstate.edu/conferences/srstw/program/session_03/ifc/index.htmlhttp://www.isip.msstate.edu/conferences/srstw/program/session_03/programming/index.htmlhttp://www.isip.msstate.edu/conferences/srstw/program/session_03/algorithms/index.htmlhttp://www.isip.msstate.edu/conferences/srstw/program/session_04/signal_processing/index.htmlhttp://www.isip.msstate.edu/conferences/srstw/program/session_05/transforms/index.htmlhttp://www.isip.msstate.edu/conferences/srstw/program/session_05/evaluations/index.htmlhttp://www.isip.msstate.edu/conferences/srstw/program/session_05/digits/index.htmlhttp://www.isip.msstate.edu/conferences/srstw/program/session_06/acoustic_modeling/index.htmlhttp://www.isip.msstate.edu/conferences/srstw/program/session_07/model_design/index.htmlhttp://www.isip.msstate.edu/conferences/srstw/program/session_07/training/index.htmlhttp://www.isip.msstate.edu/conferences/srstw/program/session_07/alphadigits/index.htmlhttp://www.isip.msstate.edu/conferences/srstw/program/session_08/language_modeling/index.htmlhttp://www.isip.msstate.edu/conferences/srstw/program/session_09/search/index.htmlhttp://www.isip.msstate.edu/conferences/srstw/program/session_09/ngrams/index.htmlhttp://www.isip.msstate.edu/conferences/srstw/program/session_09/rescoring/index.htmlhttp://www.isip.msstate.edu/conferences/srstw/program/session_10/lvcsr/index.htmlhttp://www.isip.msstate.edu/conferences/srstw/program/session_11/conversational/index.htmlhttp://www.isip.msstate.edu/conferences/srstw/program/session_11/systems/index.htmlhttp://www.isip.msstate.edu/conferences/srstw/program/session_11/switchboard/index.htmlhttp://www.isip.msstate.edu/conferences/srstw/program/session_12/internet.htmlhttp://www.isip.msstate.edu/conferences/srstw/program/session_12/participants.htmlhttp://www.isip.msstate.edu/cgi/bin/registration.plhttp://www.isip.msstate.edu/projects/speech/overview/index.htmlhttp://www.isip.msstate.edu/projects/speech/overview/index.htmlhttp://www.isip.msstate.edu/projects/speech/http://www.isip.msstate.edu/data/mailing_lists/asr/currenthttp://www.isip.msstate.edu/contact/email/http://www.isip.msstate.edu/contact/http://msuinfo.ur.msstate.edu/where/building/simrall.htmhttp://msuinfo.ur.msstate.edu/where/building/simrall.htmhttp://msuinfo.ur.msstate.edu/where/building/simrall.htmhttp://msuinfo.ur.msstate.edu/where/building/simrall.htmhttp://msuinfo.ur.msstate.edu/where/building/simrall.htmhttp://msuinfo.ur.msstate.edu/where/building/mitchell.htmhttp://msuinfo.ur.msstate.edu/where/building/mitchell.htmhttp://msuinfo.ur.msstate.edu/where/building/mitchell.htmhttp://msuinfo.ur.msstate.edu/where/building/mitchell.htmhttp://msuinfo.ur.msstate.edu/where/building/mitchell.htm

Schedule:DAY TIME SESSION DESCRIPTION PRESENTERS

SUNDAYMAY 12

15:00 -17:00

1.1: Orientation Registration,Computers

Staff

17:00 -19:00

1.2: Reception BBQ, Frisbee, softball Staff

MONDAYMAY 13

08:30 -10:00

2.1: Introduction Welcome Speech Recognition Foundation Classes

Joe Picone

10:00 -10:30

BREAK

10:30 -12:00

3.1: Classes

DSP and Templates Data Structures Algorithms, I/O, andAudio

Jie Zhao

12:00 -13:30

LUNCH (SELF-PAY)

13:30 -15:00

3.2: IFCProgramming

Math Classes Templates Data Structures

Jie Zhao

15:00 -15:30

BREAK

15:30 -17:00

3.3: Algorithms Basic DSP Buffers and I/O Audio Front-End

Jie Zhao

TUESDAYMAY 14

08:30 -10:00

4.1: SignalProcessing

Measurements Statistical Modeling TypicalImplementations

Joe Picone

10:00 -10:30

BREAK

10:30 -12:00

5.1:Transformations

Algorithms andRecipes Front-End Overview A Graphical UserInterface

Shivali Srivastava

12:00 -13:30

LUNCH (SELF-PAY)

13:30 -15:00

5.2: Evaluations

Generating Features Running Evaluations Adding NewAlgorithms

Shivali Srivastava

15:00 -15:30

BREAK



http://www.isip.msstate.edu/conferences/srstw/program/session_01/registration/index.htmlhttp://www.isip.msstate.edu/conferences/srstw/program/session_01/reception/index.htmlhttp://www.isip.msstate.edu/conferences/srstw/program/session_02/introduction/index.htmlhttp://www.isip.msstate.edu/conferences/srstw/program/session_03/ifc/index.htmlhttp://www.isip.msstate.edu/conferences/srstw/program/session_03/programming/index.htmlhttp://www.isip.msstate.edu/conferences/srstw/program/session_03/programming/index.htmlhttp://www.isip.msstate.edu/conferences/srstw/program/session_03/algorithms/index.htmlhttp://www.isip.msstate.edu/conferences/srstw/program/session_04/signal_processing/index.htmlhttp://www.isip.msstate.edu/conferences/srstw/program/session_04/signal_processing/index.htmlhttp://www.isip.msstate.edu/conferences/srstw/program/session_05/transforms/index.htmlhttp://www.isip.msstate.edu/conferences/srstw/program/session_05/evaluations/index.html

15:30 -17:00

5.3: TIDIGITS

Building DigitRecognizers Evaluating NewFront-Ends ImprovingPerformance

Shivali Srivastava

WEDNESDAYMAY 15

08:30 -10:00

6.1: AcousticModeling

Hidden MarkovModels Training Modeling Context

Mark Ordowski

10:00 -10:30

BREAK

10:30 -12:00

7.1: ModelDesign

Model Initialization Context-Independent Context-Dependent

Ram Sundaram

12:00 -13:30

LUNCH (SELF-PAY)

13:30 -15:00

7.2: Training Baum-Welch Training State Tying Mixture Generation

Ram Sundaram

15:00 -15:30

BREAK

15:30 -17:00

7.3: Alphadigits

Hierarchical Systems Lexicons andPronunciations Adding LanguageModels

Ram Sundaram

17:30 -21:00

ROAD TRIP (BRING YOUR CAMO!)

THURSDAYMAY 16

08:30 -10:00

8.1: LanguageModels

Networks and N-grams Smoothing and Pruning Search Algorithms

Joe Picone

10:00 -10:30

BREAK

10:30 -12:00

9.1: Search

Viterbi Beam Search N-gram Decoding Word Graphs andRescoring

AravindGanapath.

12:00 -13:30

LUNCH (SELF-PAY)

13:30 -15:00

9.2: N-gramModels

Generating N-gramModels Evaluating Complexity Pruning LanguageModels

Jie Zhao

15:00 -15:30

BREAK



http://www.isip.msstate.edu/conferences/srstw/program/session_05/digits/index.htmlhttp://www.isip.msstate.edu/conferences/srstw/program/session_06/acoustic_modeling/index.htmlhttp://www.isip.msstate.edu/conferences/srstw/program/session_06/acoustic_modeling/index.htmlhttp://www.isip.msstate.edu/conferences/srstw/program/session_07/model_design/index.htmlhttp://www.isip.msstate.edu/conferences/srstw/program/session_07/model_design/index.htmlhttp://www.isip.msstate.edu/conferences/srstw/program/session_07/training/index.htmlhttp://www.isip.msstate.edu/conferences/srstw/program/session_07/alphadigits/index.htmlhttp://www.isip.msstate.edu/conferences/srstw/program/session_08/language_modeling/index.htmlhttp://www.isip.msstate.edu/conferences/srstw/program/session_08/language_modeling/index.htmlhttp://www.isip.msstate.edu/conferences/srstw/program/session_09/search/index.htmlhttp://www.isip.msstate.edu/conferences/srstw/program/session_09/ngrams/index.htmlhttp://www.isip.msstate.edu/conferences/srstw/program/session_09/ngrams/index.html

15:30 -17:00

9.3: Rescoring

Graph Generation andCompaction Acoustic Rescoring Language ModelRescoring

Jie Zhao

19:00 -21:00

SRSTW'02 BASKETBALL TOURNAMENT

FRIDAYMAY 17

08:30 -10:00

10.1: LVCSRSystems

Typical LVCSRSystems Multi-Pass Systems Open Discussion

Joe Picone

10:00 -10:30

BREAK

10:30 -12:00

11.1: Real Speech Switchboard Issues in Training Efficient Decoding

AravindGanapath.

12:00 -13:30

LUNCH (SELF-PAY)

13:30 -15:00

11.2: BuildingSystems

Preparing Data Acoustic Training Word GraphGeneration

Ram Sundaram

15:00 -15:30

BREAK

15:30 -17:00

11.3: Switchboard Decoding Scoring Optimizations

Ram Sundaram

19:00 -22:00

BANQUET



http://www.isip.msstate.edu/conferences/srstw/program/session_09/rescoring/index.htmlhttp://www.isip.msstate.edu/conferences/srstw/program/session_10/lvcsr/index.htmlhttp://www.isip.msstate.edu/conferences/srstw/program/session_10/lvcsr/index.htmlhttp://www.isip.msstate.edu/conferences/srstw/program/session_11/conversational/index.htmlhttp://www.isip.msstate.edu/conferences/srstw/program/session_11/systems/index.htmlhttp://www.isip.msstate.edu/conferences/srstw/program/session_11/systems/index.htmlhttp://www.isip.msstate.edu/conferences/srstw/program/session_11/switchboard/index.html

1 - 2 - 3 - 4 - 5 - 6 - 7 - 8 - 9 - 10 - 11 - 12 - 13 - 14 - 15 - 16

Continuous Speech Recognition Using Hidden Markov Models

http://www.isip.msstate.edu/publications/journals/ieee_assp/1990/hmm/index.html (1 of 2) [3/17/2002 9:53:51 PM]

http://www.isip.msstate.edu/publications/journals/ieee_assp/1990/hmm/page_02.htmlhttp://www.isip.msstate.edu/publications/journals/ieee_assp/1990/hmm/page_03.htmlhttp://www.isip.msstate.edu/publications/journals/ieee_assp/1990/hmm/page_04.htmlhttp://www.isip.msstate.edu/publications/journals/ieee_assp/1990/hmm/page_05.htmlhttp://www.isip.msstate.edu/publications/journals/ieee_assp/1990/hmm/page_06.htmlhttp://www.isip.msstate.edu/publications/journals/ieee_assp/1990/hmm/page_07.htmlhttp://www.isip.msstate.edu/publications/journals/ieee_assp/1990/hmm/page_08.htmlhttp://www.isip.msstate.edu/publications/journals/ieee_assp/1990/hmm/page_09.htmlhttp://www.isip.msstate.edu/publications/journals/ieee_assp/1990/hmm/page_10.htmlhttp://www.isip.msstate.edu/publications/journals/ieee_assp/1990/hmm/page_11.htmlhttp://www.isip.msstate.edu/publications/journals/ieee_assp/1990/hmm/page_12.htmlhttp://www.isip.msstate.edu/publications/journals/ieee_assp/1990/hmm/page_13.htmlhttp://www.isip.msstate.edu/publications/journals/ieee_assp/1990/hmm/page_14.htmlhttp://www.isip.msstate.edu/publications/journals/ieee_assp/1990/hmm/page_15.htmlhttp://www.isip.msstate.edu/publications/journals/ieee_assp/1990/hmm/page_16.html

Up | Home | Site Map | What's New | Projects | PublicationsSpeech | Administration | About Us | Search | Contact

Please direct questions or comments to [email protected]

Continuous Speech Recognition Using Hidden Markov Models

http://www.isip.msstate.edu/publications/journals/ieee_assp/1990/hmm/index.html (2 of 2) [3/17/2002 9:53:51 PM]

http://www.isip.msstate.edu/publications/journals/ieee_assp/1990/http://www.isip.msstate.edu/index.htmlhttp://www.isip.msstate.edu/data/site_map/index.htmlhttp://www.isip.msstate.edu/whats_new/index.htmlhttp://www.isip.msstate.edu/projects/index.htmlhttp://www.isip.msstate.edu/publications/index.htmlhttp://www.isip.msstate.edu/projects/speech/index.htmlhttp://www.isip.msstate.edu/administration/index.htmlhttp://www.isip.msstate.edu/about_us/index.htmlhttp://www.isip.msstate.edu/search/index.htmlhttp://www.isip.msstate.edu/contact/index.htmlmailto:[email protected]

msstate.eduLECTURE 26: PRACTICAL ISSUES ECE 8463: FUNDAMENTALS OF SPEECH RECOGNITION LECTURE 26: PRACTICAL ISSUES SRSTW'00: SESSION 06 - ACOUSTIC MODELING LECTURE 26: PRACTICAL ISSUES LECTURE 26: PRACTICAL ISSUES LECTURE 26: PRACTICAL ISSUES LECTURE 26: PRACTICAL ISSUES LECTURE 26: PRACTICAL ISSUES LECTURE 26: PRACTICAL ISSUES LECTURE 26: PRACTICAL ISSUES SRSTW'00: TECHNICAL PROGRAM Continuous Speech Recognition Using Hidden Markov Models

LECTURE 26: PRACTICAL ISSUES · 2002. 3. 18. · LECTURE 26: PRACTICAL ISSUES Objectives: Discuss...

Documents

Transcript of LECTURE 26: PRACTICAL ISSUES · 2002. 3. 18. · LECTURE 26: PRACTICAL ISSUES Objectives: Discuss...