After successful completion of the course, a student should be able to
implement computational techniques for detection and tracking of features describing human behaviour in different modalities of perception
use machine learning methods for automatic detection of human behaviours and states
determine appropriate design approaches to build social perception abilities such as recognising humans, behaviours and higher level social states and variables (e.g., emotions) based on low-level features from different modalities of perception
describe basic principles of socially adaptive behaviour in robots and embodied interactive systems
apply basic principles of design and evaluation of human-machine interaction
evaluate the impact that affect recognition, behaviour detection, and similar technologies may have on ethical values like privacy and autonomy, and to suggest strategies for fulfilling values that are important for users and society at large, including minimisation of negative consequences
Topics include face detection and tracking, facial feature detection and tracking, facial expression and gesture recognition, automatic analysis of multimodal behaviour, automatic inference of affect and social signals, reinforcement learning for adaptive machines, machine embodiment and behaviour generation, control and planning, human-agent and human-robot interaction.
Lectures and assignments.
Written assignments (2 credits), oral and written presentation of a project (3 credits).