M.Sc. studies EMARO / Robotics
Signal Processing (ESPRO)

Winter 2017/2018

Contents:


Meeting times and locations

Lecture:
Monday, 12:15-14:00 (i.e. 0:15 p.m.- 2:00 p.m.), room 528, E&IT Faculty

Tutorial:
Monday (every second), 14:15-16:00, lab. P113, E&IT Faculty

[go to top]


Teaching staff and contact info

Prof. Włodzimierz KASPRZAK
Office: room 565, E&IT Faculty, Institute of Control and Computation Eng.
Office hours: Monday & Friday, 14.00-15.00
Phone: +22 234 7866
E-mail: W.Kasprzak (at) elka.pw.edu.pl

Maciej Stefańczyk, M.Sc.
Office: room 564
Office hours:
Phone:
Email: M.Stefanczyk (at) elka.pw.edu.pl

[go to top],


Course description

Course objectives
This course provides an introduction to digital signal processing from the perspective of audio and speech signals. Following parts are distinguished: signals and systems, Fourier Transform, digital filters, transforms and speech analysis.

Topics
Analog and digital signal conversion. Linear systems. Common signal decompositions. Convolution – its principle, properties and impulse responses. Correlation. Real and complex Fourier Transform. Applications of Fourier transform - spectral analysis of signals, frequency response of systems. Fourier transform properties. Discrete Fourier transform. Fast Fourier transform. Moving average filters. Windowed-sinc filters. Deconvolution and optimal filters. Recursive filters. Chebyshev filters. The Laplace transform and z-transform. Speech analysis.

Prerequisities
Students are expected to have the following background:

Course materials
Lecture notes will be posted periodically on the course web site. Selected chapters from the books below are recommended as optional reading.

[go to top]


Marks and Grading

Assessment will be marked out of a hundred. The marks equate to ECTS grades as given below:
ECTS Grade A B C D E F/FX
mark 100- 91 90-81 80-71 70-61 60-51 50 or less
Students are collecting assessment points. They come from a continuous assessment in the semester time: The assessment method of this course consists of: In addition to satisfying the above assessment requirements, every student must satisfy the attendance requirements. There is an obligatory attendance of tutorials and an optional attendance of lectures. The Pass mark for this course will be set at 51 pts. (for Robotics) or 60 pts. (for EMARO). Credits will be awarded to candidates who pass this course.

Lecture

Lecture schedule (tentative):


Examination tasks

  1. Part I (sec. 1-6): Test 1 (2015)
  2. Part II (sec. 7-12): Test 2 (2015)

[go to top]


Textbooks and suggested readings

Textbooks:
    Basic
  1. [1] Steven W. Smith. The Scientist and Engineer's Guide to Digital Signal Processing. Second Edition, California Technical Publishing, San Diego, CA, 1999, on-line: www.dspguide.com.
Extended
  • [2] A.V. Oppenheim, R.W. Schafer, J.R. Buck. Discrete-Time Signal Processing. Second Edition. Prentice-Hall 1999. Suggested Readings
    For each lecture section, one or more suggested readings are given below.
    (Week) Topic Readings (PDF)
    (Week 1, 2) 1. Signals Smith-ch2, Smith-ch3 : [Smith, ch. 2,3]
    (Week 3) 2. Systems Smith-ch5 [Smith, ch. 5]
    (Week 4) 3. Convolution Smith-ch6, Smith-ch7: [Smith, ch.6, 7]
    (Week 5) 4. DFT Smith-ch8, Smith-ch9, [Smith, 8,9]
    (Week 6 ) 5. FT properties Smith-ch10, Smith-ch11, [Smith, 10, 11]
    (Week 7) 6. FFT. Complex DFT Smith-ch12, Smith-ch31, [Smith, 12, 31]
    (Week 8) TEST 1
    (Week 9) 12. Speech analysis
    (Week 10) 7. Digital filters Smith-ch14, Smith-ch15, Smith-ch16, [Smith, 14,15,16]
    (Week 11) 8. Custom filters [Smith, 17,18]
    (Week 12) 9. Recursive filters [Smith, 19,20,21]
    (Week 13) 10. Laplace transform [Smith, 32]
    (Week 14) 11. z-transform [Smith, 33]
    (Week 15) TEST 2

    [go to top]


    Tutorial

    During the exercises algorithms will be discussed and implemented which illustrate the lecture material. Students will also solve their homework problems.

    Tutorial schedule

    1. E1 [9.10] : Introduction to Matlab
    2. E2 [23.10] : Signals
    3. E3 [6.11] : Systems. Convolution
    4. E4 [20.11] : Fourier Transform.
    5. E5 [4.12] : Homework assignment.
    6. E6 [18.12] :
    7. E7 [8.01] : Prototype evaluation
    8. E8 [22.01] : Final project evaluation

    Homework topics:

    Group I (project 1-3): noise detection and cancellation.
    Signal samples - 1glass.zip.

    1. Project 1 - frequency-based filtering
      Notes: SP1-SP8.

    2. Project 2 - noise estimation and subtraction
      Notes: SP-12, SP1-SP8.

    3. Project 3 - spectrogram clustering
      Notes: SP1-SP8, Introduction on clustering.

    Group II (project 4-6): auditory scene analysis.

    1. Project 4 - DOA and speaker number detection using two microphones
      Papers: DUET_2004.pdf, Kasprzak_JTIT2011.pdf.
      Signal samples - 2_two_micro_localization.zip(Matlab .mat files).

    2. Project 5 - speaker detection and localization with three microphones
      Papers: Hioka2009, Hamada_ETFA2012.pdf.
      Signal samples: 2_three_micro_localization.zip;(Matlab .mat files).

    3. Project 6 - WDO analysis of audio signals.
      Papers: DUET_2004.pdf, Kasprzak_JTIT2011.pdf.
      Audio samples: 2_localization_analysis.zip;(Matlab .mat files).

    Group III (project 7-8): audio signal separation in the time domain.

    1. Project 7 - adaptive blind source separation.
      Papers: Introduction to ICA and BSS.
      Kasprzak_2000_Ch3.pdf.
      Signal samples - 3separation.zip

    2. Project 8 - analysis of source independence.
      Papers: Introduction to ICA and BSS.
      Kasprzak_2000_Ch3.pdf.
      Signal samples - 3separation.zip

    Group IV (project 9-11): speech feature detection and analysis.
    Signal samples - 4speech.zip

    1. Project 9 - time-domain speech analysis.
      Notes: SP-12.pdf

    2. Project 10 - formant-based speech analysis.
      Notes: SP-12.pdf.
      Papers: Formant normalization and tracking.

    3. Project 11 - MFCC features of speech.
      Notes: SP-12.pdf, Introduction on clustering.
      Papers: Ch. 9 - cepstrum analysis.


    W. Kasprzak.
    Last modification: 9.10.2017