Abstract
Recent advanced in CNN acceleration systems using hardware and software co-design for real-time object detection and recognition, the speedup of post processing can reduce the burden of designing the acceleration logic to meet the real-time goal. In this paper, we propose an optimization method that eliminates the operation of acquiring redundant bounding boxes without degrading performance. It is implemented on a Xilinx Zynq Ultrascale+ MPSoC ZCU106 and shows the performance improvement about 20% that of the conventional method.
Funding Information
  • Institute for Information & communications Technology Promotion (IITP) (2019-0-01351)

This publication has 3 references indexed in Scilit: