Is it possible to train face detection, face recognition, emotion detection, voice-activity detection, and more functionality into a single CNN?
We specialize in development and research of software in the gap between current scientific study and current engineering reality.
Our main focus is software consultancy for highly technical research projects in Rust, C/C++ and Python. We design, develop and integrate the projects for customers and funders. Our expertise focuses around robotics, AR/VR, deep learning and embedded development.
In situations where the work is cutting-edge to a point where science has not yet found an acceptable answer, we do research and publish the results. Sometimes in collaboration with schools and universities, or privately for specific customers.
Next to development and research, we provide temporary management and organization services, aimed at multi-disciplinary teams with scientists, engineers and artists.
Physics simulation, robust algebra implementation, visualization and a clean parallel (Web-)GPU implementation will open up the market for state-of-the-art AR/VR and robotics apps where high-quality simulations and public visualizations overlap.
Apple has found that detecting a smile behind a face mask in times of COVID-19 is very challenging. What exactly is the problem? How much of the face does the algorithm really need to see to infer a smile? This project studies the possibilities around detecting facial expressions of people wearing face masks.
Face detection, face pose detection, eye gaze detection, visual audio detection, facial expression classification, eye target estimation, lip reading, ...
Contact us to learn more, get help, or get involved.