Core Technologies

We are developing image recognition modules using the latest advances in deep learning. Our mission is to provide developers with the building blocks for fine-grained classification of visual content. We constantly improve our image recognition technology in close collaboration with our OEM partners and vision system integrators.

Computer Vision

We use advanced digital image processing algorithms to create custom modules that allow developers to add video recognition functionalities into their software.

Read More

Deep Learning

Our Convolutional Neural Networks are trained on large-scale databases containing hundreds of millions of images and achieve exceptional accuracy.

Read More

Embedded Software

We train fast neural network models for image recognition that are suitable for embedding in edge devices. They run in real-time on CPU without sacrificing the accuracy.

Read More

Our Projects

Car Make & Model Classifier

The Vehicle MMR module recognizes 400 different vehicle brands and 7000 models, including classical cars. It recognizes car images from different views and is based on a fast neural network model with state-of-art accuracy.

Read More

Car Color Classifier

The car color classifier recognizes 14 colors: black, white, grey, silver, blue, red, green, brown, beige, golden, bordeaux, yellow, orange, and violet.

Read More

Age & Gender Classifier

The Facial Analysis module estimates age and gender from face images. The age range is from 0 to 100 years. The library is available for C++ and Python and is suitable for projects where using online APIs is not an option because of privacy concerns.

Read More