ELEC-E5500 - Speech Processing, 07.09.2020-19.10.2020
This course space end date is set to 19.10.2020 Search Courses: ELEC-E5500
Topic outline
-
Introduction to Speech Processing. Observe that the material has a slightly different organization than the lectures, so there is not always a one-to-one correspondence between lectures and chapters.Lecture materials are mainly from
Table of contents (updated as we progress):
- Introduction
- Administrative organization
- Why speech processing
- Short-time analysis
- Waveform
- Windowing
- Spectrogram
- Cepstrum
- Acoustic properties of speech
- Speech perception - the Mel scale
- Short-time processing
- Short-time Fourier Transform (STFT), filterbanks, wavelets
- PSOLA
- Short-time Fourier Transform (STFT), filterbanks, wavelets
- Classic DSP
- Time-domain processing
- Fundamental frequency (F0) estimation
- Voice activity detection (VAD)
- Speech production
- Speech production modelling (handout only)
- Speech enhancement
- Speech coding in the time domain
- Speech coding in the frequency domain
- Quality evaluation
- Speaker recognition and verification as well as speaker diarization
- Privacy and security in speech communication technology
- Quadratic problems in speech processing (omitted in 2020)
-
Video recordings of all lectures URL
-
Solution for exercise 1 is uploaded.
- Introduction