This webinar presents a complete machine learning inference flow of a convolutional neural network (CNN) architecture, including the necessary steps to implement a trained network model on an embedded target platform (Xilinx Zynq Ultrascale+ MPSoC ZCU102 development kit). This webinar is a follow-on to Accelerating Deep Learning for Vision Using Caffe Framework presented in February 2019 (link to webinar video).
• A quick recap of Artificial Intelligence applications and Deep Learning
• CMC’s machine learning training flow using Caffe Framework
• Machine learning inference flow using Xilinx DNNDK
o CNN quantization
o Hardware/software development with Xilinx SDSoC
o Deployment on a Xilinx embedded platform
• Live Demo and instructions on obtaining access to platform hardware and software