Intelligent Sensors and Integrated Systems

UNITE MNP

Extended Integrated Processor for Hardware Emulation of Evolvable Neural Spiking Systems

The project partners are expert at each level. The seed research to be developed consists in combining memristor devices for large-scale networks,
spiking neural configurable hardware and high-level brain-like models. In the light of the results, we will propose the deployment of a next-generation brain-like AI.

Objectives

Multiscale neuromorphic processing
Multiscale neuromorphic processing

  • To strengthen the collaboration between UPC, IST and KTH in research and doctoral education in the emerging area of brain-like neuromorphic computing.
  • Using a transdisciplinary perspective, we propose a new approach to conventional machine learning paradigms, applying memristive, or local adaptation, for specific recognition problems, in particular, temporal anomaly detection.
  • To prepare and submit a large-scale, ambitious research proposal in the frame of Horizon Europe, being the EIC Pathfinder 2024 program in the main focus.
  • To improve the work of the currently participant PhD students and to promote attractive thesis
    subjects to potential PhD candidates.

Partners:
UPC (coordinator), IST INESC MN, KTH

Coordinators:
Jordi Madrenas (jordi.madrenas@upc.edu)