Our paper "Spiking neurons with short-term synaptic plasticity form superior generative networks" has just been published in Nature Scientific Reports. Taking inspiration from biological synapse dynamics, we propose an important computational role of short-term plasticity in the context of pattern generation and completion. By locally modulating the energy landscape, spiking networks endowed with this mechanism can efficiently solve what is known as the mixing problem in machine learning. Find out more at https://www.nature.com/articles/s41598-018-28999-2.
Last update of this page: 2018-07-13 by Luziwei (卢子未) Leng (冷)
Electronic Vision(s) Group – Dr. Johannes Schemmel
Im Neuenheimer Feld 227
69120 Heidelberg
Germany
phone: +49 6221 549849
fax: +49 6221 549839
email: schemmel(at)kip.uni-heidelberg.de
How to find us