About the HDMP Workshop
Ground Vehicles (UGVs) have reached a crossroads: vehicles possessing a
significant level of autonomy have been demonstrated yet those vehicles
still cannot be safely and reliably utilized in populated environments. At the same time, the
automotive industry is under increasing pressure to reduce the number
of accident related fatalities and has embarked upon research to
develop technical solutions including intelligent driver warning
systems. Hence, human detection has
attracted many researchers for over a decade.
The ability to accurately detect,
and project the future trajectory of people represents a major barrier
to the large scale employment of UGVs for both civil and military
applications. The nature of these applications requires both very high
true pedestrian detection rate and very low false detection rate.
practical aspects of the problem must also
be addressed -- such as real-time performance
constraints. Solving this problem is critically important in both
military and civilian applications.
vision and robotics researchers are invited to submit original
contributions in the following -- but not limited to
detection in bad weather condition
humans in non-upright postures
detection using different modalities (such as Infrared, LADAR and RADAR)
human detection by fusing different modalities.
scale estimation mechanisms
tracking from mobile platforms