Neuronal controllers for cognitive robots

In our group “Neuromorphic Cognitive Systems", we develop neuronal architectures that allow robots to generate behavior (e.g. navigate in an environment, avoid obstacles, pursue targets), to form memories (e.g. build a map), and learn. We realise these architectures in neuromorphic hardware and strive to achieve efficiency of biological neuronal systems. In a student project, different contributions to this goal can be made, including (1) neuronal SLAM (simultaneous localisation and mapping); (2) object detection and classification in a navigation scenario (using a neuromorphic camera); (3) reinforcement learning; (4) sequence learning. Depending on the interest, experience, skills, motivation, and the available time of the student, projects with different scope and difficulty can be defined.

Requirements

Programming experience or wish to learn (C/C++, Python, Matlab, or Java); Interest in neuronal systems (biological or artificial); interest in AI, robotics, or cognitive science.

Contact

Yulia Sandamirskaya: yulia.sandamirskaya(at)ini.uzh.ch

© 2017 Institut für Neuroinformatik