Simulating visual spiking neural networks using real life experimental data is HARD, but it's one of the main objectives of SOLIS!
#neuralnetwork #STDP #SRDP #simulation

The complete pipeline of SOLIS: from experimental characterisation of synapses to visual computing

Here is what a SOLIS #synaptic network can do for #visualcomputing.
A cycle of memroy writting and progressive erasing using Kesterite-based synapses. Patterns and edge detection are the first step for a proof of concept of #neuromorphic visual computing!
The full code will be released next year.

#artificial_synapses are cool, but true visual #neural_networks are cooler.
SOLIS is creating, characterising, and simulating true hardware neural networks based on #thin_films.
Here, we show how a 3x3 #synaptic_network can adjust its weights and recognise a pattern. The code will be shared soon.

We published a new version of our software to control a Keithley series 26xx (dual channel). This allows to characterise synapses, transistors, and solar cells in a really intuitive way. Check it out!
zenodo.org/records/1739...

#keithley #sourcemeter #solarcell #transistor #synapse

Webpage: cordis.europa.eu/project/id/1...
GitHub: github.com/SOLIS-project
Zenodo: zenodo.org/communities/...

Consortium:
Polytechnic University of Catalonia (Spain)
University of Liverpool (UK)
University of Verona (Italy)
Tallinn University of Technology (Estonia)
UP Diliman (Philippines)
University of Western Cape (South Africa)
Moroccan Foundation for Advanced Science Innovation and Research (Morocco)

Hello.
SOLIS is a EU-funded project based on staff and knowledge exchange. It revolves around the thematic of artificial visual synapses based on thin film materials and inspired by photvoltaic solar cells.
The project will run until March 2029 and involved 8 international institutions!