The neuromorphic approach to iCub

Lecture
Datum: 
18. April 2017
Beginn: 
11:30
Ende: 
12:30
Raum: 
CITEC, 1.204

Robotics and AI hold the promise of permeating our daily lives in the near future. More specifically, robots of humanoid form can interact naturally with people as well as reuse the tools and environments already designed for human presence. Improving their perception and ability to react quickly to environmental stimuli is paramount for autonomy, needed for the effective deployment of robots in this scenario. This research line is about the design of cutting-edge, efficient technology for robot perception. This is obtained by combining hardware design and, consequently, by developing special algorithms for data processing. In particular, hardware is in the form of neuromorphic electronics for event-driven encoding of sensory signals, algorithms in the form of event-based signal processing.