SnapML: How to Run Machine Learning in Snapchat

Snapchat, the popular Social app, launched SnapML last June: an important update to its development tool (Lens Studio) that allows the use of Machine Learning algorithms to create Lens, that is filters that enrich the user experience .

Snapchat is not new to this type of initiative having always been a pioneer in the adoption of innovative tools or new technologies, just remember for example the Spectacles: glasses on which to use Augmented Reality applications (2016).

Photo credits — Wareable.com

More recently, October 2020, they have included in Lens Studio, full support for the iPhone 12 Pro Max LIDAR (which I hope to cover in another article).

Photo Credits — Engadget

SnapML

SnapML is the name Snapchat gave to support for Machine Learning algorithms. Broadly speaking, the development model provides:

1. the creation and training of a neural network in an external tool (ie Tensorflow or Pytorch)
2. inserting the model created in a Lens Studio project
3. testing the filter, Lente in SnapChat language, with your Smartphone
4. Distribution of the created Lens to SnapChat users