Sidebar Menu

Lecture "Pattern Recognition and Machine Learning"

 

Basic Information

 
 
Lecture overview  
  Lecturers:   Gerhard Schmidt (lecture) and Tobias Hübschen, Bastian Kaulen (exercise)  
  Room:   Zoom  
  Language:   English  
  Target group:   Students in electrical engineering and computer engineering  
  Prerequisites:   Basics in system theory  
  Contents:  

In this lecture the basics of speech, audio, and music signal processing are treated. Often schemes that are based on statistical optimization are utilized for these applications. The involved cost function are matched to the human audio perception.

Topic overview:

  • Preprocessing to reduce signal distortions
    • Noise reduction
    • Beamforming
  • Feature extraction and data compression
  • Pattern recognition and data regression
    • Codebooks
    • Gaussian mixture models (GMMs)
    • Artificial Neural Networks (ANNs)
    • Hidden Markov models (HMMs)
  • Selected applications of machine learning
 

 

News

A preliminary lecture and exercise schedule for WS 20/21 is now available. The lecture will be conducted as a live Zoom meeting, scheduled every Tuesday, 13:00 h. The first lecture starts on November, 3rd. The exercise will be provided as on-demand videos. There will also be two Zoom sessions for the excercise where students can ask questions regarding both the lecture and the exercise.

The Zoom link for the lecture has been distributed to the students registered in OLAT. If you did not receive a link, please contact This email address is being protected from spambots. You need JavaScript enabled to view it..

 

Schedule

The following schedule regarding lectures and excercises is preliminary and may be adapted during the semester. The lecture will usually take place from 13:00 h - 15:30 h.

 
 
Date   Lecture   Exercise  
  03.11.2020   Introduction   -  
  10.11.2020   Noise Suppression + Beamforming   Noise Suppression (video)  
  17.11.2020   Beamforming + Feature Extractiom   Beamforming (video)  
  24.11.2020   Feature Extraction + Codebook Training   Feature Extraction (video)  
  01.12.2020   Codebook Training + Bandwidth Extension   Codebook Training (video)  
  08.12.2020   Bandwidth Extension (until 15:00 h)   Zoom Question Time (starts 15:15 h, topics up to Codebook Training) + Bandwidth Extension (video)  
  15.12.2020   Gaussian Mixture Models   Gaussian Mixture Models (video)  
  05.01.2021   Neural Networks   Neural Networks (video)  
  12.01.2021   Hidden Markov Models   Hidden Markov Models (video)  
  19.01.2021   Speaker and Speech Recognition   Speaker and Speech Recognition (video)  
  26.01.2021   -   Zoom Question Time (starts 13:00 h, all topics)  
  02.02.2021   Student Talks   Student talks  
  09.02.2021   Student Talks   Student talks  

 

Lecture Slides

 
 
Link   Content  
    Slides of the lecture "Introduction"
(Introduction, boundary conditions of the lecture, applications)
 
    Slides of the lecture "Noise Suppression"
(Noise suppression, dereverberation, speech reconstruction)
 
    Slides of the lecture "Beamforming"
(Fixed and adaptive beamforming, postfiltering)
 
    Slides of the lecture "Feature Extraction"
(Linear prediction, cepstrum, mel-filtered cepstral coefficients)
 
    Slides of the lecture "Codebook Training"
(K-means algorithm, LBG algorithms)
 
    Slides of the lecture "Bandwidth Extension"
(Model-bases approaches, evaluation)
 
    Slides of the lecture "Gaussian Mixture Models (GMMs)"
(Training with the EM algorithm, applications)
 
    Slides of the lecture "Hidden Markov Models (HMMs)"
(Efficient probability calculation, training of HMMs)
 
    Slides of the lecture "Speaker and Speech Recognition"
(Application of GMMs and HMMs, speech dialog systems)
 
    Slides of the lecture "Neural networks"
(Network types, training procecures)
 

 

Matlab Demos

 
 
Link   Content  
    Matlab demo (GUI based) for adaptive noise suppression  
    Matlab demo (GUI based) for linear prediction  

 

Exercises

The exercise will consist of two parts using different media. For most lecture topics an individual exercise video will be uploaded. These videos can be watched on demand. Some additional course material will be provided alongside these videos.

The on-demand videos will be supported with two live zoom sessions. Here, students will be able to ask topic-related questions. Students are also encouraged to submit their questions ahead of time so the answers can be supported with slides or other material.

Videos will go live and zoom sessions will be conducted according to the (preliminary) schedule above.

 
 
Video   Content   Material  
   

Noise suppression:

  • Wiener Filter summary
  • comprehension questions
  • python demo
 

 
   

Beamforming:

  • beamforming summary
  • comprehension questions
  • python demo
 

 
   

Feature extraction:

  • linear prediction
  • LPC, LPCC, MFCC features
  • comprehension questions
  • python demo
 

 

 

Talks

At the end of the semester, each student will give a talk about a certain topic. The aim is both to give you the chance to work on a pattern recognition-related topic that interests you, and to improve your presentational skills. The talk is also a prerequisite for your admission to the exam. The talks should be held in English and should take ten minutes, plus 2.5 minutes of discussion and 2.5 minutes of feedback. Please write an email to This email address is being protected from spambots. You need JavaScript enabled to view it. to reserve your topic. The talk registration deadline is 24.01.2021.

Below you can find the schedule of the talks.

 
 
Date   Room   Time   Topic   Presenter(s)  
                     

 

 

 

Evaluations

 
 
Link   Content   Link   Content  
    Current evaluation     Completed evaluations