synnexgrp/openvino repository overview

⁠SYNNEX Intel AI Sales Kit

SYNNEX Intel AI Sales Kit pre-install Intel OpenVINO Toolkits, it helps you to run the demos and benchmark your device in a easy way. You can use SYNNEX Intel AI Sales Kit to

• Run Benchmark of Yolo_v3 and Yolo_v4 model
• Run Smart City Demos
• Run and Download more demos and AI models by including OpenVINO_Demo_Kit⁠

In next sections, we will guide you to set up environment, pull docker image and run SYNNEX Intel AI Sales Kit.

⁠Getting Start

⁠Installation

⁠STEP 1 : Install OS

⁠Download and Install Ubuntu 20.04.2.0 LTS

HERE⁠ to donwload Ubuntu 20.04.2.0 LTS ISO image

⁠Install Ubuntu 20.04

HERE⁠ is a guide for Ubuntu installation.

⁠STEP 2 : Install Docker

sudo apt update
sudo apt-get update
sudo apt-get install docker.io

Inorder to run as non-root user, you have to run commands below.

sudo usermod -aG docker $USER

If you still have some problem about permission. Try with

sudo chmod 777 /var/run/docker.sock

⁠STEP 3 : Pull Docker image

• ⁠For OpenVINO 2021.4 on TigerLake

docker pull synnexgrp/openvino:2021.4_tgl

• ⁠For OpenVINO 2021.4

docker pull synnexgrp/openvino:2021.4_developer_models

• ⁠For OpenVINO 2021.3

docker pull synnexgrp/openvino:2021.3_developer_models

⁠Run Demo App and Tool

⁠1. Smart City Demo

# Download the demo video
wget -O ~/Downloads/NewVideo2.mp4 https://github.com/incluit/OpenVino-For-SmartCity/raw/master/data/NewVideo2.mp4
# Run demo in the docker image with the video and show the result
docker run -it -v /tmp/.X11-unix:/tmp/.X11-unix -e DISPLAY=$DISPLAY -v ~/Downloads:/mnt --device /dev/dri:/dev/dri --group-add=$(stat -c "%g" /dev/dri/render*) --rm synnexgrp/openvino:2021.4_developer_models /home/openvino/smartcity_demo.sh

• ⁠Please check your image tag is match as you want to run. ("2021.4_developer_models" or "2021.3_developer_models")

⁠2. Run Human Pose Estimation Demo

The Demo Video is a web source, need internet connection to run this demo.

# Run demo in the docker image with the video and show the result
docker run -it -v /tmp/.X11-unix:/tmp/.X11-unix -e DISPLAY=$DISPLAY -v ~/Downloads:/mnt --device /dev/dri:/dev/dri --group-add=$(stat -c "%g" /dev/dri/render*) --rm synnexgrp/openvino:2021.4_developer_models /home/openvino/human_pose_estimation.sh

This is a demo using OpenVINO DL Streamer,demonstrates human pose estimation pipeline constructed via gst-launch-1.0 command-line utility. For more information, please goto OpenVINO Online Documents⁠.

⁠3. Run Vehicle and Pedestrian Tracking Demo

The Demo Video is a web source, need internet connection to run this demo.

# Run demo in the docker image with the video and show the result
docker run -it -v /tmp/.X11-unix:/tmp/.X11-unix -e DISPLAY=$DISPLAY -v ~/Downloads:/mnt --device /dev/dri:/dev/dri --group-add=$(stat -c "%g" /dev/dri/render*) --rm synnexgrp/openvino:2021.4_developer_models /home/openvino/vehicle_pedestrian_tracking.sh

This is a demo using OpenVINO DL Streamer, demonstrates gvatrack⁠ element and object tracking capabilities on example of person and vehicle tracking. Object tracking increases performance by running inference on object detection and classification models less frequently (not every frame). For more information, please goto OpenVINO Online Documents⁠.

⁠4. Run Benchmark of Yolo_v3 and Yolo_v4 model

docker run -it -v /tmp/.X11-unix:/tmp/.X11-unix -e DISPLAY=$DISPLAY -v ~/Downloads:/mnt --device /dev/dri:/dev/dri --group-add=$(stat -c "%g" /dev/dri/render*) --rm synnexgrp/openvino:2021.4_developer_models

• ⁠Please check your image tag is match as you want to run. ("2021.4_developer_models" or "2021.3_developer_models")

Then you will get into the docker, run following command to excute the benchmark test.

cd
python3 run_command.py

⁠Run OpenVINO_Demo_Kit

SYNNEX Intel AI Sales Kit includes OpenVINO_Demo_Kit, which is a tool that can make you run intel openVINO Demos and samples easily. It helps you to install openVINO, build demos, download/convert models,run DLWorkbench automatically. No need to run those demos/samples or operations manully with long arguments and path. Also, this Demo Kits support to run benchmark app with multiple models dump the throughputs, Latency,etc. in one .csv file. It might help you to suvery your device's AI performance on different kinds of model. Here's the demo that can support with AI Sales Kit.

• Security Barrier Camera Demo
• Interactive Face Detection Demo
• Crossroad Camera Demo
• Image Processing Demo
• Instance Segmentation Demo
• Image Segmentation Demo
• Text Spotting Demo
• Action Recognition Demo
• Gesture Recognition Demo
• Face Recognition Demo
• Social Distance Demo
• Whiteboard Inpainting Demo

⁠Optional step for camera

If you want to run demo with camera, please use following command to run the AI Sales Kit.

docker run --device=/dev/video0:/dev/video0 -it -v /tmp/.X11-unix:/tmp/.X11-unix -e DISPLAY=$DISPLAY -v ~/Downloads:/mnt --device /dev/dri:/dev/dri --group-add=$(stat -c "%g" /dev/dri/render*) --rm synnexgrp/openvino:2021.4_developer_models

• ⁠Please check your image tag is match as you want to run. ("2021.4_developer_models" or "2021.3_developer_models")

You can replace the video ID with yours. Also can change the folder to mount to /mnt if you want. Then goto OpenVINO_Demo_Kit folder to run the Demo_Kit

cd OpenVINO_Demo_Kit
./Demo_Kit.sh

For more detail message and user guides, please go to OpenVINO_Demo_Kit Intro Website⁠ . Or pull an issue⁠

⁠Support

Please visit SYNNEX website⁠ or contact [email protected]⁠ for more information about Intel AI Solutions.