Gordon Cooper, Product Marketing Manager for Embedded Vision Processors at Synopsys, demonstrates the company's latest embedded vision technologies and products at the May 2017 Embedded Vision Summit. Specifically, Cooper shows highly accurate object classification using the VGG16 neural network running on the DesignWare EV6x Vision Processor (on a HAPS FPGA-based prototyping platform). The EV6x includes a scalar and vector DSP as well as a convolutional neural network (CNN) engine that classifies images from Caffe. The demo shows the excellent results of the trained graph.
