Introduction
This project will introduce a technology field growing exponentially in popularity in Machine Learning and Artificial Intelligence – TinyML. Due to the depth of this field, this project has been broken down into four articles, which go into the nitty gritty details about this topic eventually.
What will we build?
We will first discuss Tiny ML and its differences from the existing Machine Learning methodologies. Then you will be introduced to a trendy platform used for developing TinyML models, Edge Impulse. Next, we will get familiar with the hardware devices we will be using throughout this series: Arduino Nano 33 BLE Sense and Sony Spresense, which are two popular IoT development boards. Finally, we will create our ver first Hello World! Project wherein we will control the RGB LED of our Arduino BLE sense using simple voice commands. The model will be trained and deployed using the Edge Impulse Platform.
Then we will build three prototype level projects using the Arduino BLE Sense and the Sony Spresense board.
-
The first project will cover the creation of a project in which we will be building a TinyML model using collected Audio data to add voice control capabilities to existing smart home projects.
-
In the second project, we will add face recognition capabilities to build a complex Home security system using TinyML and the Sony Spresense and camera board.
-
Finally,in the third project we will build a gesture recognition model, which will allow our Arduino BLE Sense to perform different tasks based on detected gestures. It will be like building your very own magin wand!
For all the detailed tutorials, do checkout the full articles for this project on Circuit Cellar.