Arduino and ml5.js

Machine learning frameworks require advanced knowledge of calculus, linear algebra, statistics, computer science, etc. While important for research and development of new machine learning models and architectures, starting from this point is very discouraging. In the summer of 2017, deeplearn.js paved the way for making machine learning easier to access by bringing it to the web. Deeplearn.js was an open source hardware-accelerated javascript library and didn’t require any environment to be installed and configured. In 2018 deeplearn.js became TensorFlow.js and enabled to use Tensorflow framework directly in the browser.

In our course, we will use ml5.js which provides an easy-to-use access to machine learning algorithms and models in the browser based on Google’s TensorFlow.js. In ml5.js you can either build your network from scratch or use a variety of pre-trained models which are based on Image, Sound and Text inputs.