Setup Visual Studio Code environment for Vision ML

What you will do

  • Install a Visual Studio Code (VS Code) dev environment for Windows, Linux or MacOS
  • Setup and Configure VS Code for container deployment
  • Clone Vision AI Dev Kit GitHub repository using VS Code sample

What you will need

  • VS Code
  • Python 3.7 by Anaconda
  • Vision AI DevKit extension for VS Code
  • Azure account with configured Azure IoT Hub
  • Azure user account with configured Azure IoT Hub

Windows users

Setup and Configure VS Code for container deployment

  1. If not already available, install VS Code (Download) and Python 3.7 by Anaconda (using default options). (Download)

  2. Once installed, open the Anaconda Prompt from Start Menu\Programs\Anaconda3 (64-bit)

  3. Open VS Code. Click the Extensions icon in the Activity Bar (left side) and install Vision AI Dev Kit extension. It bundles together several other needed extensions: VS Code extensions

  4. Install the Azure ML SDK (Instructions) and required packages.

    • Note: When installing the Azure Machine Learning SDK or related Python packages, there are some Python packages which depend on specific versions of other Python packages. If these (dependency) Python packages versions are too high, the installation will fail. To work around this issue, there is a requirements.txt in GitHub under VisionSample\MachineLearning\scripts folder that can be used to specify correct versions. Please copy the file in the VisionAIDevKit root folder (ex. c:\visionaidevkit\requirements.txt) which has stricter versions defined. Below is the content of the file if you prefer creating it locally.

    Recommended requirements.txt contents:

  5. Use the Anaconda Prompt (base) as an administrator, then execute the following command lines:

     pip install --upgrade -r requirements.txt
  6. Restart Visual Studio Code in the Python 3.7 runtime environment, by executing the following command in an Anaconda Prompt (base) window:


    This will enable VS code to inherit the environment variables of the Python 3.7 runtime environment.

  7. Open the command palette in VS Code (Ctrl-Shift-P), then enter

     Python: Select Interpreter

    to select your Python interpreter (Python 3.7)

  8. In the VS Code command pallette, enter

     Azure: Sign In

    to sign in to your Azure account and select your subscription.

  9. Create a new workspace in VS Code as described in Get started with Azure Machine Learning for Visual Studio Code. You can also use script described in the next section to create a new workspace.

Linux and MacOS users

Linux/iOS - these instructions are in draft form and are still being improved

For Linux these have been tested for Ubuntu or Ubuntu VM version 18.04.

Install and configure required tools

  1. Install the build tools.

    sudo apt-get install build-essential checkinstall
    sudo apt-get install libreadline-gplv2-dev libncursesw5-dev libssl-dev
       \ libsqlite3-dev tk-dev libgdbm-dev libc6-dev libbz2-dev
    sudo apt-get install libffi-dev
  2. Download Python 3.7.2.

    cd /usr/src
    sudo wget
  3. Extract tgz

    sudo tar xzf Python-3.7.2.tgz
  4. Compile.

    cd Python-3.7.2
    sudo ./configure --enable-optimizations
    sudo make altinstall
  5. Add alias into ~/.bashrc

    alias python='/usr/local/bin/python3.7m'
  6. Install pip:

     curl -o
     sudo python
     sudo pip install --upgrade pip

Setup and Configure VS Code for container deployment

  1. Install VS Code (Download Here)

  2. Open VS Code. Click the Extensions icon in the Activity Bar (left side) search and install the Vision AI DevKit extension:
  3. Install the Azure ML SDK (Instructions) and required packages.

Clone GitHub

  1. Open command palette (“View” -> “Command palette”)
    • Select “Azure IoT Edge Show Sample Gallery”
    • Click Vision AI Developer Kit -> Open Sample
    • Select local folder to use to store the sample solution

Open sample

GitHub view in VS Code