课程信息

Course Contents
Algorithms of speech and audio signal processing: Introduction to the models of speech and audio signals and basic methods of audio signal processing. Procedures of codebook based processing and audio coding. Beamforming for spatial filtering and noise reduction for spectral filtering. Cepstral filtering and fundamental frequency estimation. Mel-filterind cepstral coefficients (MFCCs) as basis for speaker detection and speech recognition. Classification methods based on GMM (Gaussian mixture models) and speech recognition with HMM (Hidden markov models). Introduction to the methods of music signal processing, e.g. Shazam-App or beat detection.

Literature
Slides (for further details see homepage of the lecture)

Preconditions
Knowlegde about satistical signal processing is required (lecture „Digital Signal Processing“). Desired
– but not mandatory – is knowledge about adaptive filters.

Online Offerings
Moodle