Module CS5274-KP08

Duration

1 Semester

Turnus of offer

each summer semester

Credit points

8

Course of studies, specific fields and terms:

• Master Auditory Technology 2022, compulsory, Auditory Technology
• Master Auditory Technology 2017, compulsory, Auditory Technology

Classes and lectures:

• Speech and Audio Signal Processing (lecture, 2 SWS)
• Selected Topics of Signal Analysis and Enhancement (lecture, 2 SWS)
• Selected Topics of Signal Analysis and Enhancement (exercise, 1 SWS)
• Speech and Audio Signal Processing (exercise, 1 SWS)

Workload:

• 40 hours exam preparation
• 90 hours in-classroom work
• 110 hours private studies

Contents of teaching:

• Speech production and human hearing
• Physical models of the auditory System
• Dynamic compression
• Spectral analysis: Spectrum and cepstrum
• Spectral perception and masking
• Vocal tract models
• Linear prediction
• Coding in time and frequency domains
• Speech synthesis
• Noise reduction and echo compensation
• Source localization and spatial reproduction
• Basics of automatic speech recognition
• Introduction to statistical signal analysis
• Autocorrelation and spectral estimation
• Linear estimators
• Linear optimal filters
• Adaptive filters
• Multichannel signal processing, beamforming, and source separation
• Compressed sensing
• Basic concepts of multirate signal processing
• Nonlinear signal processing algorithms
• Application scenarios in auditory technology, enhancement, and restauration of one- and higher-dimensional signals, Sound-field measurement, noise reduction, deconvolution (listening-room compensation), inpainting

Qualification-goals/Competencies:

• Students are able to describe the basics of human speech production and the corresponding mathematical models.
• They are able to describe the process of human auditory perception and the corresponding signal processing tools for mimicing auditory perception.
• They are able to present basic knowledge of statistical speech modeling and automatic speech recognition.
• They can describe and use signal processing methods for source separation and room-acoustic measurements.
• Students are able to explain the basic elements of stochastic signal processing and optimum filtering.
• They are able to describe and apply linear estimation theory.
• Students are able to describe the concepts of adaptive signal processing.
• They are able to describe and apply the concepts of multichannel signal processing.
• They are able to describe the concept of compressed sensing.
• They are able to analyze and design multirate systems.
• Students are able to explain various applications of nonlinear and adaptive signal processing.
• They are able to create and implement linear optimum filters and nonlinear signal enhancement techniques on their own.

Grading through:

• Written or oral exam as announced by the examiner

Responsible for this module:

• Prof. Dr.-Ing. Alfred Mertins

Teacher:

• Institute for Signal Processing
• Prof. Dr.-Ing. Alfred Mertins

Literature:

• L. Rabiner, B.-H. Juang : Fundamentals of Speech Recognition Upper Saddle River: Prentice Hall 1993
• J. O. Heller, J. L. Hansen, J. G. Proakis : Discrete-Time Processing of Speech Signals IEEE Press
• A. Mertins : Signaltheorie: Grundlagen der Signalbeschreibung, Filterbänke, Wavelets, Zeit-Frequenz-Analyse, Parameter- und Signalschätzung Springer-Vieweg, 3. Auflage, 2013
• S. Haykin : Adaptive Filter Theory Prentice Hall, 1995

Language:

• German and English skills required

Notes:

Last Updated:

17.08.2022