Research

Alongside advancing the theory of spiking neural networks, our research team is also heavily engaged in applied research to leverage neuromorphic AI algorithms for real-world applications. The outcomes of these studies are proof-of-concept (PoC) software and hardware systems that demonstrate the advantages of neuromorphic solutions and evaluate their potential performance levels across a variety of application domains.

Currently, the applied research is focused on the key properties of neuromorphic solutions such as energy efficiency, high performance, security (resistance to adversarial attacks), and adaptivity (the ability to fine-tune a neural network with a small number of examples during operation).