Future Memcomputing Arrays for Next Generation Computer Vision

Project Description

The project will develop a new type of optical memristor (OM) that will be integrated within a novel memcomputing architecture
based on a cellular nonlinear network (CNN). This leverages the unique capacity of OMs to detect light and store
information, while also exploiting CNNs ability to simultaneously process the information in all cells at once. This highly
parallel, in-memory computing engine will enable ultra-fast, real-time processing of visual information (~30,000
frames/second), which has applicability to a wide range of current societal needs (safer autonomous driving, better security
systems) and numerous applications in medicine.