Fundamentals of Speech Recognition
As with most ICISLT courses, this
course is a hands-on project-based course. However,
this course covers subject matter similar to a number of
conventional graduate courses in speech recognition so there
are textbooks available for reference. There will be no
lectures, however, and the textbook material can be studied
in a just-in-time fashion according to the needs of each
project.
Textbooks:
Frederick Jelinek,
Statistical Methods for Speech Recognition
Xuedong Huang
et. al., Spoken Language Processing
Projects:
Preliminary Projects
-
Reading and Writing Shared File Formats
-
Generating Data from a Hidden Markov Process
-
Generating Unit Tests for Projects
Analysis
Tools
Acoustic
Modeling
-
Alignment to a Script
-
Training Acoustic Models
Language
Modeling
Search Methodologies
Recognition Systems
|