UNIVERSITY OF HERTFORDSHIRE
	     COMPUTER SCIENCE RESEARCH COLLOQUIUM

			    presents


	"Visual Navigation with a Tiny Brain: Getting Home
		  Without Knowing Where You Are"

		      Dr. Andrew Philippides
	  (Centre for Computational Neuroscience and Robotics,
		       University of Sussex)

		     22 November 2017 (Wednesday)

			   1 - 2 pm


		  Hatfield, College Lane Campus
			Seminar Room C408

		  Everyone is Welcome to Attend

		  Refreshments will be available


Abstract:

The use of visual information for navigation is a universal strategy for
sighted animals, amongst whom social insects are particular experts. The
general interest in studies of insect navigation is in part due to their
small brains; biomimetic engineers can take inspiration from elegant and
parsimonious control solutions, while biologists look for a description of
the minimal cognitive requirements for complex spatial behaviours. We take
an interdisciplinary approach to studying visual guided navigation. By
combining behavioural experiments with modelling and robotics we show
how animals directly acquire and use task-specific information through
specialized sensors, brains and behaviours, enabling complex behaviour
to emerge without complex processing.

In this spirit, I will show that an agent – insect or robot – can
robustly navigate without ever knowing where it is, without specifying
when or what it should learn, nor requiring it to recognize specific
objects, places routes or maps. This leads to an algorithm in which
navigation is driven by familiarity detection rather than explicit recall,
with sensory data specifying actions not locations. Route navigation
is thus recast as a search for familiar views, allowing an agent to
encode routes through visually complex worlds in a single layer neural
network after a single training run with only a low resolution panoramic
sensor. Combined, this means that computational requirements in terms
of memory and processing are low making the algorithm suitable for small
robots, UAVs and wearable sensors.


About the Speaker:

Dr. Andrew Phillipides is Reader in Computer Science and Artificial
Intelligence, Centre for Computational Neuroscience and Robotics,
University of Sussex, and co-director of the Centre for Computational
Neuroscience and Robotics (http://www.sussex.ac.uk/ccnr/). Having read
Mathematics at King’s College Cambridge, he moved to Sussex  to do
an MSc in Artificial Intelligence and Adaptive Systems followed by a
doctorate in Computational Neuroscience and Robotics, and has been at
Sussex since. His research is interdisciplinary and is best described
as computational neuroethology; That is, it combines behavioural
experiments with computational and robotic models to understand, and take
inspiration from, biological systems. His current research topics include:
visual navigation in insects and robots, neuromodulation in (real and
artificial) neural networks, analysis of biological imaging data and
agent-based modelling applied to crowd movement and human migration.



---------------------------------------------------
Hertfordshire Computer Science Research Colloquium
http://cs-colloq.cs.herts.ac.uk