We have more and more devices at home, these can perform several actions, but they need some intelligence to do that.
Home assistant allows us to automate tasks in our home by providing a hub where you can connect devices and give them instructions. It also allows some "if then, do that" logic, however, we felt like a more powerful brain was still missing.
During the last week, I created a small Pipeless Agents integration for Home Assistant, which allows you to automate your home using computer vision by connecting your existing cameras, no need to change them.
Until now, we could perform basic actions thanks to motion sensors and basic people recognition, but with this approach, we can allow our home to make its own decisions on what to do.
I created a tutorial so you can understand the basics. On it, you will learn how to connect Home Assistant with Pipeless Agents, set up a project, add your camera streams, and implement your custom logic and video filters.
We will deploy a simple example that turns off our TV when the people watching it leave the scene. You can continue playing around with that basic implementation and create more complex workflows and applications.
You don't need to know anything about computer vision since your code receives a structured data stream, you just need very basic Python knowledge.
The following is the complete step-by-step tutorial:
Top comments (0)