Informatik 6 - Advanced Automatic Speech Recognition, SS 2010

Advanced Automatic Speech Recognition, SS 2010

News - Schedule - Download - Exercises - Info - Literature - Contact

News

April 20, 2010

New Lecture: Advanced Automatic Speech Recognition

In this semester, for the first time we offer an advanced lecture on automatic speech recognition. This lecture builds on the Lecture Introduction to Automatic Speech Recognition which is a requirement for the attendance of this advanced lecture. This lecture addresses students in master courses as well as doctoral studies.

Contents:

recapitulation of automatic speech recognition with large vocabulary and linear lexicon
search with lexical prefix trees
word graphs and word pair approximation
time-conditioned search
coarticulation
confidence measures and system combination
speaker adaption and normalization
discriminative training
current topics

Schedule

Type	Date	Room	Start	Instructor/s
V3	Tue, 14:00 - 15:30 (biweekly)	Seminarraum LS Informatik 6 (Raum 6124)	April 20, 2010	Dr.rer.nat. R. Schlüter, Prof. Dr.-Ing. H. Ney
	Thu, 14:00 - 15:30	Seminarraum LS Informatik 6 (Raum 6124)	April 22, 2010
Ü1	Tue, 14:00 - 15:30 (biweekly)	Seminarraum LS Informatik 6 (Raum 6124)	April 27, 2010	D. Nolden, Dr.rer.nat. R. Schlüter

Specific schedule (individual distribution of lectures/exercises):

Month	Day	Lecture/Exercise
April	20	Lecture
	22	Lecture
	27	Exercises
	29	Lecture
May	4	Lecture
	6	Lecture
	11	Exercises
	13	Holiday: Ascension Day
	18	Lecture
	20	Lecture
	25	Holiday: Excursion Week/Pentecost
	27	Holiday: Excursion Week/Pentecost
June	1	DIES: RWTH Sports Day
	3	Holiday: Corpus Christi
	8	Lecture
	10	Exercises (Thursday!)
	15	Lecture
	17	Lecture
	22	Exercises
	24	Lecture
	29	Lecture
July	1	Lecture
	6	Exercises
	8	Lecture
	13	Lecture
	15	Lecture
	20	Exercises
	22	Lecture
	27	Lecture
August	5	Lecture

Download

Here you can find the download material for the lecture. The exercise sheets as well as other documents needed for the solution of the exercises will be made available on this here.

Lecture Notes:

Final version of lecture notes of this lecture (Advanced Automatic Speech Recognition) (Chapter 6 [MLLR/fMLLR] reworked)

Lecture notes of previous Lecture Introduction to Automatic Speech Recognition

The RWTH Aachen University Speech Recognition System can be downloaded here.

Exercises

Exercises:

The exercises should be treated in groups of 3 students.
Requirement for the issuing of certificates (Schein/Leistungsnachweis) and for the admission to the exam will be that at least 50% of the exercise points are obtained.
Solutions to problems without points will be presented in the exercise lesson.
Exercise sheets will be provided at the exercise lesson 2 preceding the submission date, i.e. two weeks in advance.

1^st Exercise Sheet (not to be submitted - will be presented/explained on April 27^th, 2010)

2^nd Exercise Sheet (Submission: May 11^th, 2010)

3^th Exercise Sheet (Submission: June 10^th, 2010)
4^th Exercise Sheet (Submission: June 22^th, 2010)
5^th Exercise Sheet (Submission: July 6^th, 2010)
6^th Exercise Sheet (Submission: July 20^th, 2010)

Info

Content:
The aim of automatic speech recognition is to extract the spoken word sequence from the acoustic signal. An automatic speech recognition system consists typically of four parts:

the signal analysis extracts time-dependent features from the acoustic signal
the acoustic model establishes a link between features and phonemes
the language model covers syntactic and semantic constraints of the language
the search process determines the word sequence that fits best to the features taking the acoustic and language model into account.

Assignment:

Master of Science Informatik
Diplom Informatik, Applied Computer Science or Field of Specialization
Master Software Systems Engineering
Master Media Informatics

Requirements:

Introduction to Automatic Speech Recognition

Language of instruction:

English

References

Emphasis on signal processing and small-vocabulary recognition: L. Rabiner, B. H. Juang: Fundamentals of Speech Recognition, Prentice Hall, Englewood Cliffs, New Jersey, 1993.

L. Rabiner, R. Schafer: "Fundamentals of Speech Recognition", Foundations and Trends® in Signal Processing, Vol. 1, No 1-2, pp. 1-194, 2007.

(free download)

Emphasis on large vocabulary and language modelling: F. Jelinek: Statistical Methods for Speech Recognition, MIT Press, Cambridge, 1997.
Introduction to both speech and language: D. Jurafsky, J. H. Martin: Speech and Language Processing, Prentice Hall, Englewood Cliffs, NJ, 2000.
Advanced topics: R. De Mori: Spoken Dialogues with Computers Academic Press, London, 1998.
M.Gales, S. Young: "The Application of Hidden Markov Models in Speech Recognition", Foundations and Trends® in Signal Processing, Vol. 1, No 3, pp. 195-304, 2008. (free download)

Contact

Dr. Ralf Schlüter	schlueter@informatik.rwth-aachen.de	Tel. 80-21612
David Nolden	nolden@i6.informatik.rwth-aachen.de	Tel. 80-21602

Office hours: by arrangement