OY ripaoj 2 testi2 EnLaajuus (5 cr)

Objective

After completing the course, student 1. is able to explain the emotion theory and modeling 2. is able to implement algorithms for emotion recognition from visual and audio signals or the fusion of multi-modalities 3. has the ideas of wide applications of affective computing

Content

The history and evolution of affective computing; psychological study about emotion theory and modeling; emotion recognition from different modalities: facial expression, speech, EEG; crowdsourcing study; synthesis of emotional behaviors; emotion applications

Qualifications

Basic mathematical skills from bachelor's and the courses Data structures and algorithms and Introduction to artificial intelligence or equivalent. Programming skills using Python (Python 3.6). Advanced courses such as Machine learning and Computer vision are also beneficial but not required.

Assessment criteria, satisfactory (1)

The assessment of the course is based on the exam (30%) and a research report (70%) with mandatory exercises. A passing grade from both the exam and research report is required.

Materials

Lecture slides, pre-collected codes, tools and databases.

Execution methods

The course consists of eight lectures and three exercises. In the lectures the teacher introduces basic knowledge and theories. In the exercises, the teacher and TAs will tutor students to work on hands-on tasks of progamming and implementing affective computing algorithms.

Accomplishment methods

Students have to complete three mandatory parts to get the final grade: 1. Three programming exercises 2. One research report 3. Final exam. The assessment of the course is based on the exam (30%) and a research report (70%) with mandatory programming exercises. A passing grade from both the exam and research report is required.