Intel® FPGA AI Suite: PCIe-based Design Example User Guide

ID 768977
Date 4/05/2023
Public

A newer version of this document is available. Customers should click here to go to the newest version.

Document Table of Contents

5.7.1. Example Running the Object Detection Demonstration Application

You must download the below items:
  • yolo-v3-tf from the OpenVINO™ Model Downloader. The command should look similar to below.
python3
        <path_to_installation>/open_model_zoo/tools/downloader/downloader.py \
   --name yolo-v3-tf \
   --output_dir <download_dir>
  • Next generate .bin/.xml from the downloaded model.
python3
        <path_to_installation>/open_model_zoo/tools/downloader/converter.py \
   --name yolo-v3-tf \
   --download_dir <download_dir> \
   --output_dir <output_dir> \
   --mo <path_to_installation>/model_optimizer/mo.py
        

To run the object detection demonstration application, execute the following (assuming the FPGA has been configured with the Generic bitstream):

./runtime/ object_detection_demo_yolov3_async/
object_detection_demo_yolov3_async \
   -d HETERO:FPGA,CPU \
   -i <path_to_input_video.mp4> \
   -m <path_to_generated_graph.xml> \
   -plugins_xml_file ./runtime/plugins.xml \  
   -arch_file=./example_architectures/A10_Generic.arch \
   -t 0.65