Embedded World 2017: Autonomous Perception Demo Using Deep Learning, Dense Optical Flow and Stereo Vision - Presented by Xilinx

This autonomous drive demonstration runs the three major complex algorithms commonly used in autonomous systems today, Convolutional Neural Network (CNN), Dense Optical Flow and Stereo Vision, all in a single Zynq Ultrascale+ MPSoC device. The entire design was developed using the new Xilinx embedded vision stack allowing developers to directly take Caffe trained network and OpenCV functions and integrate with sensors / displays using only C/C++.