DEEP OSCILLATORY NEURAL NETWORKS COMPUTING AND LEARNING THROUGH THE DYNAMICS OF RF NEURONS INTERCONNECTED BY RF SPINTRONIC SYNAPSES
Developing artificial neural networks that can learn, process and 'think' as the brain does is a sort of Holy Grail with virtually limitless applications. Deep learning paradigms exploiting multilayer hierarchical artificial neural networks mimicking the brain's structure can also mimic the brain's ability to learn by example. They have achieved tremendous success, even exceeding human performance in some cases. The EU-funded RadioSpin project plans to demonstrate deep learning networks processing radiofrequency (RF) signals and learning at speeds up to 10 million times faster than a human brain. Benchmarking applications will target mammography and IoT RF fingerprinting.
Total Eligible Budget
INL Eligible Budget
Type of action
RIA Research and Innovation action
Grant Agreement Id
FET – Future Emerging Technologies