FaceBlur

Scenario

Detects faces into pictures using the Azure Cognitive Services Face API and blurs them to anonymize the information.

Original Picture | Final Result :-------------------------:|:-------------------------: |

What's in this repo

In this repository you'll find different ways and languages to deploy and solve the problem!

First of all, if you need some test images you can find them here. These images have been downloaded from unsplash.

You can deploy the solution in two different ways:

• Using a Jupyter Notebook and Azure Machine Learning Service (how-to here)
• Using Azure Functions (how-to here). In the functions folder you can find the deployment both in python and C#, so choose the language you prefer!

Before you start

You can run the solution locally, using Visual Studio Code and few extensions, Visual Studio or you can run it on Azure. Depending on which option you'll choose you need to set up the environment.

Visual Studio Code

Download Visual Studio Code from free here and install it on your device. Once you have installed it, go to Extensions (Ctrl+shift+X if you want to use shortcuts) and install the following extensions:

• Azure Account
• Azure Functions
• Python
• Jupyter
• C#

Visual Studio

todo

Azure subscription

Navigate to the Azure Portal and login in to your subscription. If you don't have one, you can create your Azure free account here.

Let's try it!

Notebook

To deploy and run the notebook follow the instruction you find here

Functions

Architecture

Functions languages

You can deploy and run the function both in C# and python. Choose the language you prefer and follow the instruction:

• C# function
• python function

Resources

• Face API
• What is the Azure Face Service
• Detect faces in an image using Python
• Azure Functions
• Azure Functions docs
• Azure Function with Python
• Azure Functions - Blob trigger