|
|
|||||||||||||||||
Course Catalog 2012-2013
SGN-4106 Speech Recognition, 5 cr |
Additional information
Suitable for postgraduate studies
Person responsible
Tuomas Virtanen, Jani Nurminen, Annamaria Mesaros
Lessons
| Study type | P1 | P2 | P3 | P4 | Summer | Implementations | Lecture times and places |
|
|
|
|
|
|
|
|
|
Requirements
Final examination and exercises.
Principles and baselines related to teaching and learning
-
Learning outcomes
After completing this course, the student will understand the basic techniques used in speech recognition. He or she will be able to implement the front-end used for extracting relevant information from the speech signal. The student will have a detailed understanding of the mathematical principles of hidden Markov models (HMMs) that are used to model the data provided courtesy of the front-end. The student will be able to calculate quantities that are requred to train HMMs and use them for pattern classification. The student will be able to implement a simple HMM-based speech recognition system.
Content
| Content | Core content | Complementary knowledge | Specialist knowledge |
| 1. | Front-end of a speech recognizer, cepstral coefficients, phonetics | ||
| 2. | Hidden Markov models: training the models and using them for pattern classification. | ||
| 3. | The use of hidden Markov models for automatic speech recognition. | ||
| 4. | Language models |
Evaluation criteria for the course
Exam and exercises. Excellent grade can be obtained by answering to the questions in an exam by the extent the topic have been dealth with on the lectures and exercises. Approximately half of the maximum number of points need to be obtained and 20% of the exercises need to be completed in order to pass the course.
Assessment scale:
Numerical evaluation scale (1-5) will be used on the course
Study material
| Type | Name | Author | ISBN | URL | Edition, availability, ... | Examination material | Language |
| Lecture slides | Bryan Pellom | English |
Prerequisites
| Course | Mandatory/Advisable | Description |
| SGN-2500 Introduction to Pattern Recognition | Advisable | 1 |
| SGN-2506 Introduction to Pattern Recognition | Advisable | 1 |
| SGN-4010 Speech Processing Methods | Mandatory |
1 . Prior knowledge about pattern recognition is advisable.
Additional information about prerequisites
SGN-4010 Speech Processing or corresponding knowledge of speech processing is required.
Prerequisite relations (Requires logging in to POP)
Correspondence of content
| Course | Corresponds course | Description |
|
|
|
|
|
|
|
More precise information per implementation
| Implementation | Description | Methods of instruction | Implementation |