BMVA
British Machine Vision Association and Society for Pattern Recognition
Call for Participation
Vision Systems for Perception and Action
One Day BMVA technical meeting in London, UK, on 6th March, 2009
Chairs: Dr Neil Thacker, University of Manchester,
Dr E. Charles Leek, Bangor University
The task of constructing artificial vision systems still holds many
challenges, many of which are shown into sharp relief by the apparent
ease with which humans solve visual processing tasks. By comparison,
many computer vision algorithms have very restricted applicability,
particularly with respect to natural scenes. There appears to be more to
be learned regarding which information is used and at what point during
human visual processing it is obtained. What should be regarded as a
semantically useful description of the world? Which tasks are inevitably
`top down', and which should be achievable `bottom up'?
Great strides have been made within the last twenty years in our
understanding of the cognitive and neurobiological substrates of visual
perception and motor control. These advances have begun to unravel the
complex inner workings of the human brain providing us with unique
insights into the powerful computational systems that have evolved to
support seemingly effortless activities like perceiving shape, reaching
and grasping.
The purpose of this one day BMVA meeting is to bring together
researchers from areas of biological modelling, psychophysics and
computer vision, to discuss the strengths and weaknesses of the
computational models used when attempting tasks equivalent to human
visual perception. It is intended that as well as covering material
describing latest results in understanding human vision there will be
more general talks which summarise the experiences and opinions of
representatives from these disciplines, including challenges to current
approaches. We will investigate how our understanding of human vision
and motor control may be exploited by computer vision researchers, and
provide constraints on the development of more efficient artificial
vision and robotics systems. In turn, it is also recognised that
advances in computer vision themselves provide further insights into
human cognitive processing.
Time will be allocated for discussion of the issues raised so that the
possibilities for consensus and future synergies between these fields
can be explored. If you have views on this subject please attend and
join in with the discussions.
Although the agenda has been organised by invitation some space is still
available for additional speakers which fit in with the objectives of
the meeting. If you wish to discuss the possibility of making a
presentation please send an email to [log in to unmask] by
6th February 2009.
|