Novel oscillatory neural network technology for improved human-machine interactions and nanoelectronics.
Our world has never been more digitized. As AI applications and nanoelectronics continue to grow, the use of sensors is important. Currently, most sensors receive analogue inputs from the real world and generate analogue signals to be processed. The inevitable digitization of these signals, however, will create enormous amounts of raw data requiring a lot of memory and high power consumption. As the number of sensor-based IoTs grows, bandwidth limitations will also need to be addressed. In this context, the EU-funded PHASTRAC project will develop an analogue-to-information neuromorphic computing paradigm based on oscillatory neural networks (ONNs). The aim of this ONN computing architecture will be to seamlessly interface with sensors and process their analogue data without any analogue-to-digital conversion.