SSD is a widely used one-stage object detection model suitable for many AI applications such as video surveillance and ADAS.

We recently published a comprehensive tutorial that walks users through all the steps to run the SSD model on a ZCU102 platform. 

You can access this tutorial here:

It is designed to help developers understand how to use Xilinx’s DNNDK tools to quantize, compile and deploy a Caffe-trained SSD model on Xilinx SoC platforms. The main tutorial steps are shown below:

We will continue to add high-quality Edge AI tutorials on the Xilinx Github.