course: Fundamentals of Speech Signals Processing

number:
141165
teaching methods:
lecture with tutorials
media:
handouts, computer based presentation, black board and chalk
responsible person:
Prof. Dr.-Ing. Rainer Martin
lecturer:
Prof. Dr.-Ing. Rainer Martin (ETIT)
language:
german
HWS:
3
CP:
4
offered in:
winter term

dates in winter term

  • start: Tuesday the 08.10.2019
  • lecture Tuesdays: from 10:15 to 11.45 o'clock in ID 03/419
  • tutorial Tuesdays: from 16:15 to 17.45 o'clock in ID 03/419

Exam

All statements pertaining to examination modalities (for the summer/winter term of 2020) are given with reservations. Changes due to new requirements from the university will be announced as soon as possible.

Date according to prior agreement with lecturer.

Form of exam:oral
Registration for exam:FlexNow
Duration:30min

goals

The participants master the basic terms and models of speech signal processing and are able to use them in the context of contemporary applications. They are able to explain the generation of an acoustic speech signal and its properties in all essential details in temporal and spectral domains. They know the components and characteristics of speech coding methods and their application in mobile and packet-switched telephony as well as the methods of noise reduction, as they are used in smartphones and hearing aids.

content

This lecture deals with the basic principles and methods of digital speech signal processing, as they are used among others in telephony, mobile telephony, hearing aids, and in speech-controlled human-machine interfaces. Here, the digital model of speech production and applications in speech transmission are at the centre of attention. In particular, it is dealt with the following topics:

  • The source-filter model of speech production
  • The characteristics of the speech signal in time, frequency and cepstral domains
  • Linear prediction
  • Quantisation of scalar and vector quantities
  • Speech coding and speech transmission
  • Packetised speech transmission (Voice-over-IP, Push-to-talk)
  • Adaptive filters for noise reduction

In the exercises and computer exercises (some as homework) selected questions are investigated in greater depth..

requirements

keine

recommended knowledge

  • Systemtheorie
  • Grundkenntnisse der digitalen Signalverarbeitung,
  • Grundkenntnisse der Wahrscheinlichkeitsrechnung und der stochastischen Signale

literature

  1. Martin, Rainer, Vary, Peter "Digital Speech Transmission. Enhancement, Coding and Error Concealment", Wiley & Sons, 2006

miscellaneous

Vorlesung und Übungen werden, falls erforderlich, in englischer Sprache abgehalten. Die Lehrmaterialien stehen in Deutsch und in Englisch zur Verfügung.

Die einzelnen Übungstermine werden in der Vorlesung bekannt gegeben. Die Note setzt sich aus den Ergebnissen der mündlichen Prüfung und der studienbegleitenden Leistungen (Hörsaalübungen, Rechnerübungen) zusammen. Genaueres ist den Vorlesungsunterlagen (Moodle) zu entnehmen. Im Fall einer Wiederholung der Prüfung im Sommersemester, wird die Note allein aus der mündlichen Prüfung gebildet.