Unlocking a new interface
with Deep Learning
Our solution can understand complex hand-drawings from business and technology contexts. It detects and classifies drawn objects, their relationships, dependencies, hierarchies, and the broader intention and expression of the user creating the sketch.
This makes sketched information available for a variety of different applications and opens up a way of communication, that was previously inaccessible to machines.
Sparse Visual Intelligence
We analyse highly sparse imagery through custom developed machine learning methods. Our models can perform recognition, object detection and segmentation tasks on images that were previously impossible to process.
In the age of deep learning, OCR needs to step up its game. Our systems strives to compete on a human level. They repair, heal and cognify corrupted or obfuscated document elements in order to make the information available digitally.