Experiments with Speech Recognition.
| Sponsor |
Irfan Essa
irfan@cc.gatech.edu
320 CoC
Chris Atkeson
cga@cc.gatech.edu
CRB 378
|
| Area |
Intelligent Systems / GVU |
Problem
Speech Recognition systems are getting much commercial attention these
days and many "dictation" type systems are proving to be
quite successful. This mini-project is aimed at introducing you to some
of the important concepts related to speech recognition and also to
experiment with one such system.
Here is what you need to do.
-
Read the following classic paper on speech recognition. This will give
a very good review of why speech recognition is important.
-
Reddy, R., "Speech Recognition by Machine: A Review", In IEEE
Proceedings 6(4), 502--531, 1976.
-
Read the following paper, which gives an excellent review of methods
used in speech recognition.
-
Rabiner, L., "A Tutorial on Hidden Markov Models and Selected Applications
in Speech Recognition", In Proceedings of the IEEE, 77(2),
257--286, 1989.
-
Study the HTK and Waves system by Entropics Inc. which is
toolkit for speech recognition available in the Computational
Perception Lab.(One of my students will show you how).
-
Develop a small demo of speech recognition.
-
Use other systems we have available at the College for Speech
Recognition.
-
Evaluate all the systems and discuss the differences, strengths,
weaknesses.
Background
Deliverables
-
Write a 3-5 page Report addressing the following issues:
-
Give a brief overview of what you understand is the speech recognition
problem. Is it tractable?
-
What is the main point of Reddy paper.
-
Explain the hidden Markov Models
-
Report your progress with the HTK/waves package.
-
Test out other available systems
-
Compare all the systems you tried out. discuss the differences,
strengths, weaknesses.
Evaluation
Based on the report turned in to the sponsor of the project by the due
date.