Jay Young

I am a Research Fellow in Intelligent Robotics, part of the Intelligent Robotics group, in the School of Computer Science at the University of Birmingham, United Kingdom. I work as part of both the STRANDS and ALOOF projects, where we are interested in helping Mobile Robots better understand the objects and places they encounter in the world. I am broadly interested in these things:

  • Integrated AI systems deployed long-term in unprepared, real-world environments.
  • Evolutionary/Genetic Computation and Optimisation.
  • Motivation frameworks, goal generation and drive-based systems.
  • Computer Vision, perception pipelines, object recognition and semantic object/environment mapping.
  • Machine Learning on Robots over long time-spans, especially evolutionary learning to optimise goal generation and drive prioritisation.

I currently work with our robots Bob and Betty, where I am responsible for developing and maintaining our Object Learning pipelines. This means I work on the intersection of Computer Vision, planning, learning, representation, data processing, system architecture and a bunch of other things to build learning systems and produce object-centric Robot behaviour. On the ALOOF project we are working to bridge the gap between the wealth of information found on the web and a robot's situated perception and experiences. One way we do this is by using object recognition algorithms trained on very large, ever-expanding web-based image databases, to enable a robot to be able to recognise hundreds or thousands of objects, where traditionally they may be limited to a few dozen, which have to be manually taught. We also use mining of the semantic web to extract information about objects and places, so robots can acquire a more semantically rich representation of their environment.

Contact


Jay Young
School of Computer Science
University of Birmingham
Edgbaston
Birmingham, B15 2TT
E-mail: j.young@cs.bham.ac.uk