A unique way to think about algorithms.
Deep learning has brought successful performance and tuning automation to algorithms facing high data complexity and variability.
We blend explicit features all over them to leverage these advantages while decreasing their footprint and black box effects.
A focus on computer vision.
Our expertise is tailored to computer vision applications, building on their geometric specificities and expressiveness.
A fresh take on frugality, beyond TinyML.
Finely intertwining neural networks with specific functions helps us produce smaller algorithms with less data.
to know if and how our tech can help you.