440-4224/01 – Speech Processing (ZŘS)
| Gurantor department | Department of Telecommunications | Credits | 4 |
| Subject guarantor | Ing. Jaromír Továrek, Ph.D. | Subject version guarantor | Ing. Jaromír Továrek, Ph.D. |
| Study level | undergraduate or graduate | Requirement | Optional |
| Year | 2 | Semester | winter |
| | Study language | Czech |
| Year of introduction | 2016/2017 | Year of cancellation | |
| Intended for the faculties | FEI | Intended for study types | Follow-up Master |
Subject aims expressed by acquired skills and competences
After completing the course, students will be able to solve problems in the field of speech processing. They will learn the basic approaches and methods of speech signal processing, such as feature extraction and processing by neural networks or hidden Markov models. They master to implement a simple system to identify the speaker or the recognition of emotion from speech signal.
Teaching methods
Lectures
Tutorials
Experimental work in labs
Summary
Area of speech processing is one of the important part of information and communication technology. The goal of the course is to understand of basic tasks of speech processing which are SI (Speaker Identification), ASR (Automatic Speech Recognition), TTS (Text to Speech) and SER (Speech Emotion Recognition). Acquired skills can be used for design complex systems where the speech processing is used.
Compulsory literature:
Recommended literature:
Way of continuous check of knowledge in the course of semester
Test (0-15) points
Project (0-25) points
E-learning
http://lms.vsb.cz/
Other requirements
No additional requirements are placed on the student.
Prerequisities
Subject has no prerequisities.
Co-requisities
Subject has no co-requisities.
Subject syllabus:
Subject syllabus
1. Introduction to subject and speech processing, practical applications and its using.
2. Speech production, basic concepts, speech preprocessing (DC Offset, preemphases, segmentation, windowing).
3. Basic features - energy, zero cross ratio (ZCR), Jitter, Shimmer, autocorrelation, fundamental frequency.
4. Spectrum, spectrogram, spectral analysis of vowels and consonants.
5. Cepstrum, cepstral analysis, Mel frequency cepstral coefficients and other speech parameters.
6. Introduction to classification, SOM, k-NN, GMM, ANN and classifier fusion.
7. Speaker identification (SI) and possible approaches.
8. Speech emotion recognition (SER), stress recognition.
9. Automatic speech recognition (ASR) and possible approaches.
10. Text to speech (TTS), speech corpora and open-source projects.
Excercise syllabus
1. Introduction, Safety, Conditions for subject completion.
2. Practical exercises – speech preprocessing – DC offset, preemphases, segmentation, windowing.
3. Practical exercises – Feautures extraction – energy, zero cross ratio, fundamental frequency.
4. Practical exercises – Spectral analysis of speech signal.
5. Practical exercises – Features extraction – MFCC, LPC.
6. Test and assigment of project.
7. Design of speaker recognition system - GMM, ANN.
8. Example of project proposal.
9. Speech synthesis.
10. Presentation of projects.
Conditions for subject completion
Occurrence in study plans
Occurrence in special blocks
Assessment of instruction
Předmět neobsahuje žádné hodnocení.