NGSLT at KTH
October 31 – November 1, 2005

Speech Technology Level 1

Fantum, TMH, KTH, Stockholm

Program

Monday, October 31

Speech and Speaker Recognition

10.15 – 11.00 Speech Recognition 1: Acoustic Analysis, Modelling, Search
Kjell Elenius

11.15 – 12.00 Speech Recognition 2: Training, Adaptation

Mats Blomberg

12.00 – 13.15 Lunch

13.15 – 13.45 Finite State Transducers

Alec Seward

13.45 – 14.15 Dialog system demonstration

Gabriel Skantze

14.30 – 15.00 Speaker Verification

Mats Blomberg

15.00 – 15.30 Coffee break

15.30 – 18.00 Speech Recognition Laboration

Alec Seward

18.30 Pizza

TMH Lunch room

Tuesday, November 1

Multimodal Speech Synthesis

9.00 – 11.00 Multimodal speech synthesis - Lecture in Fantum

Björn Granström

11.00 – 11.30 Discussion of the phonetic analysis assignments - Fantum

David House

11.30 – 12.30 Lunch

12.30 – 14.00 Speech synthesis practical – Basement computer room

14.00 – 14.30 Presentation/discussion of results from practical - Basement computer room

14.30 – 15.00 Coffee break

15.00 Seminar by Håkan Melin in Fantum (optional)
“Distribution of speaker verification errors in time and (speaker) space”