For Edge (DPUCZDX8G) - 1.4 English

Vitis AI User Guide (UG1414)

Document ID
Release Date
1.4 English
  1. Download the vitis_ai_runtime_r1.2.0_image_video.tar.gz from host to the target using scp with the following command.
    $scp vitis_ai_runtime_r1.2.0_image_video.tar.gz root@[IP_OF_BOARD]:~/
  2. Unzip the vitis_ai_runtime_r1.2.0_image_video.tar.gz package.
    #tar -xzvf vitis_ai_runtime_r1.2.0_image_video.tar.gz -C ~/Vitis-AI/VART
  3. Enter the directory of samples in the target board. Takeresnet50 as an example.
    #cd ~/Vitis-AI/VART/samples/resnet50
  4. Run the example.
    #./resnet50 model_dir_for_zcu102/resnet50.elf
    Note: If the above executable program does not exist, you have to cross-compiler it on the host first.
    Note: For examples with video input, only `webm` and `raw` format are supported by default with the official system image. If you want to support video data in other formats, you need to install the relevant packages on the system.

The following table shows the run commands for all the Vitis AI samples.

Table 1. Launching Commands for Vitis AI Samples on ZCU102
ID Example Name Command
1 resnet50 ./resnet50 model_dir_for_zcu102/resnet50.elf
2 resnet50_mt_py python3 1 model_dir_for_zcu102/resnet50.elf
3 inception_v1_mt_py

python3 1 model_dir_for_zcu102/inception_v1_tf.elf

4 pose_detection ./pose_detection video/pose.webm model_dir_for_zcu102/pose_0/sp_net.elf model_dir_for_zcu102/ssd/ssd_pedestrain_pruned_0_97.elf
5 video_analysis ./video_analysis video/structure.webm model_dir_for_zcu102/ssd_traffic_pruned_0_9.elf
6 adas_detection ./adas_detection video/adas.webm model_dir_for_zcu102/yolov3_adas_pruned_0_9.elf
7 segmentation ./segmentation video/traffic.webm model_dir_for_zcu102/fpn.elf