Automatic Speech Recognition

Lecture Notes

A full set of lecture slides is listed below, including guest lectures. Lectures 3, 4, and 6 have audio links to speech samples presented during the lectures.

WEEK #

LEC #

TOPICS

1

1
2

Course Overview (PDF)
Acoustic Theory of Speech Production
 (PDF - 1.4 MB)

2

3
4

Speech Sounds (PDF - 3.6 MB)
Speech Sounds (continued)

3

5
6

Signal Representation (PDF - 1.9 MB)
Vector Quantization
 (PDF - 1.8 MB)

4

7
8

Pattern Classification (1) (PDF - 1.1 MB)
Pattern Classification (2) (PDF)

5

9
10

Search (PDF)
Hidden Markov Modeling (1) (PDF)

6

11
12

Language Modeling (PDF)
Language Modeling (continued)

7

13

 

Guest Lecture by Karen Livescu: Graphical Models (PDF)
Quiz 1

8

14
15

Guest Lecture by Rita Singh: Hidden Markov Modeling (2) (PDF - 2.1 MB)
Guest Lecture by Rita Singh: Hidden Markov Modeling (3) (PDF - 1.4 MB)

9

16
17

Segment-Based ASR (PDF)
Guest Lecture by Lee Hetherington: Finite-State Transducers (PDF)

10

18
19

Acoustic-Phonetic Modeling (PDF)
Robust ASR (1) (PDF)

11

20
21

Guest Lecture by Timothy Hazen: Robust ASR (2) (PDF)
Guest Lecture by Timothy Hazen: Adaptation (PDF)

12

22
23

Speech Understanding (PDF - 1.1 MB)
Guest Lecture by Timothy Hazen: Paralinguistic Information (PDF - 1.0 MB)

13

 

Quiz 2
No Lecture

14

 

Term Project Presentations

Assignments

ASSN #

TOPICS

SUPPORTING MATERIALS

1

Acoustic Theory (PDF)

An Introduction to LAMINAR (PDF)
An Introduction to Using WAVES+ (PDF)

2

Speech Sounds (PDF)

3

Signal Representation (PDF)

4

Acoustic Modeling (PDF)

5

Hidden Markov Models 1 (PDF)

Supplement to Q9 and Q10 (PDF)

6

Language Modeling (PDF)

7

Graphical Models (PDF)

Suggested Readings (PDF)

8

Hidden Markov Models 2 (PDF)

9

Segment-Based ASR (PDF)

Assignment Errata (PDF)