This unit will give students a foundation in the subject of cognitive robotics and machine vision. For the cognitive robotics part, this will involve an introduction to cognitive robotics and the integration of machine learning methods for robots’ cognitive architectures. It will also focus on methods and algorithms for human-robot interaction and social robots and language and speech interfaces to communicate with robots. For the computer vision part, this will involve gaining familiarity with algorithms for low-level and intermediate-level processing and considering the organisation of practical systems. Particular emphasis will be placed on the importance of representation in making explicit prior knowledge, control strategy and interpreting hypotheses. This course unit treats vision as a process of inference from noisy and uncertain data and emphasizes probabilistic and statistical approaches.
Topics covered in the course include: Introduction to cognitive robotics; Developmental, evolutionary and swarm robotics; Human-robot interaction and social robots; Language and speech interfaces; Deep learning; Introduction to computer vision; Visual object recognition and tracking; Vision-based robot localisation and navigation; 3-D human and hand pose estimation; Motion generation using learnt computational models of human motion.
This course unit is designed for students that are interested in Cognitive Robotics and Human-Robot Interaction, Computer Vision, Artificial Intelligence, or Machine Learning. This course unit is also appropriate for students with an interest in Computer Graphics and/or Robotics.
The unit will consist of interactive lectures and labs (computer vision and machine learning software labs and robot demos).
The assessment for this course unit is based on a combination of coursework and a closed-book exam. The coursework consists of: reports on a set of practical assignments carried out using using robotics and computer vision libraries. Feedback will be provided via Blackboard.