Work-efficient parallel non-maximum suppression for embedded GPU architectures

International Conferences 2016