Alberto Patiño Saucedo, postdoctoral researcher at the Instituto de Microelectrónica de Sevilla (IMSE-CNM), has been awarded the best paper by the Neural Systems and Applications Technical Committee of the IEEE Circuits and Systems Society (CAS). He presented at this year's ISCAS Congress in Singapore.
The article is titled Co-optimized training of models with synaptic delays for digital neuromorphic accelerators. This is a new method to train and deploy spiking neural networks (SNNs) on digital neuromorphic hardware (that is, it imitates the functioning of the human brain).
This advance not only optimizes synaptic weights, but also synaptic delays, thus achieving better performance, since it allows artificial intelligence models to be deployed much more efficiently, reducing memory and energy consumption per synaptic operation. Being a case of edge switching (edge computing), It opens the door to faster and safer more advanced technologies, and could have applications in fields such as medicine, robotics and consumer technology.
From the Instituto de Microelectrónica de Sevilla we want to congratulate you on this international award.
You can expand the information of the paper in the following link.
Instituto de Microelectrónica de Sevilla
June 10, 2024