Yi-Wen Liu

EE6641 Analysis and Synthesis of Audio Signals  Spring semester, 2013


Class Time: Monday 13:20-15:10, Thursday 09:00-09:50
Class Location: Delta Rm 201

Instructor: Prof. Yi-Wen Liu       ywliu_(at)_ee.nthu.edu.tw

Instructor Office Hour: Thursdays 10:30-11:30; 14:30-15:30.

Teaching Assistants: 吳致暐
TA Hours: To be announced.

Course Announcements:

  • Welcome! This year the course will be organized into Monday lectures and Thursday group discussion.
  • Please check course materials weekly.  All handouts, slides, and homework assignments will be posted on NTHU LMS system.
  • Note to NCTU students: Please consider creating an account on NTHU's LMS e-learning system (數位學習系統). Otherwise, we should discuss alternative ways of distributing assigned reading materials.

Grading policy:

  • Homework assignments (40%): There will be four homework assignments, all involving MATLAB programming.
  • Quizzes (30%): There will be quizzes. The purpose is to make sure we understand the basics.
  • Final project (20%): 1st-year MS students are encouraged to come up with individual research proposal and obtain pilot results by the end
    of the semester. Otherwise, you can choose to do a hands-on team project.
  • Class participation (13%): 10% attendance on Thursdays and 3% bonus points, potentially to lift your score across grade boundaries.
  • Summing up all the above, your final grade will be converted to letter base (A+,A,A-,etc).

Syllabus:

Week # Date Monday Syllabus
(Course materials)
Thurday Group Discussion Deadlines and evaluations
1 2/18,21 Lecturer off-campus for an international conference; the TA will explain course logistics. If not familiar with MATLAB, please consider attending a MATLAB tutorial given by my lab members on 2/21 16:30 at Delta Rm 217.  
2 2/25 Part I: Fourier Transform and applications
Discrete-time Fourier Transforms; Linear-time invariant systems and convolution theorems
[National Holiday, class will not meet]  
3 3/4,7 Short-time Fourier transforms and time-frequency dualities Allen and Berkley (1979) "Image method for room-acoustics simulation"  
4 3/11,14 Overlap-add system for sound analysis and synthesis Oppenheim (1970) "Speech spectrogram using FFT" HW1 Due: Room-acoustics simulation
5 3/18,21 Overlap-add system (cont'd)
Part II: Music
Spectral Estimation I: overview
Oppenheim and Monasco (2013) "Human acuity beats the uncertainty principle"

 

6 3/25,28 Quiz #1
Spectral Estimation II: pitch estimation
Rocchesso et al. (2003). "Sounding objects" Quiz #1
7 4/1 Parametric audio coding: the sine + noise decomposition [Spring break]  
8 4/8,11 Part III: Speech
Introduction to Phonetics and Phonology
-- HW2 Due: Pitch estimation and FIR filter implementation
9 4/15,18 Speech analysis and synthesis 經典書籍導讀: S. Pinker "the Language Instinct  
10 4/22,25 The Levinson-Durbin algorithm Wee et al. (2011) "An Articulatory Silicon Vocal Tract for Speech and Hearing Prostheses"  
11 4/29,5/2 (open discussion for final projects) Gilbert and Feng (2008) "Speech and Language Processing over the Web"
HW3 Due: Sinusoidal modeling
12 5/6, 9 Part IV: Psychology, physiology, and medical applications
Psychoacoustics and perceptual audio coding

(Quiz #2)

Information hiding [tentative] Quiz #2
Final project proposal
13 5/13,16 The auditory systems and biomimetic filterbanks Lyon (2011). Google's attempt to analyze all YouTube sound tracks  
14 5/20,23 Remedies for hearing impairments
[invited speaker:
馬偕醫學院聽語系 李沛群 教授]
Cochlear implant [tentative] HW4 Due: Voice analysis and synthesis
15 5/27,30 Computational Auditory Scene Analysis TBD  
16 6/3,6 (Quiz #3) [Lecturer off-campus, class will not meet.] Quiz #3
Final-project manuscript
17 6/10,13

ICASSP/ICA post-conference sharing

TBD  
18 6/17

Final Presentation

--- Revised final paper