[2023] Monostable multivibrator networks for low-cost spike-based computing

Neuromorphic engineering aims at replicating the brain's key computational primitives in silicon, with the end goal of solving real-world AI tasks in a more efficient way than conventional machine-learning approaches. However, since a neuron is a living cell first and only then a computational unit, it is not clear whether replicating biological neurons in silicon is the best path toward an optimal solution.

In this project, we want to reverse the question. Instead of asking how we can mimic a biological neuron in hardware, we will investigate which existing elementary electronic building blocks show interesting dynamical and computational properties for artificial neurons, regardless of their biological plausibility. A first step in that direction consists in exploring monostable multivibrators (MMVs) as novel artificial neurons [1]. MMVs are simple timers that can be scaled up easily as counters in digital hardware (FPGA/ASIC). They form a class of non-biologically inspired spiking neurons. Networks of MMVs can be trained with state-of-the-art spike-based training approaches [2] and exhibit interesting dynamical behavior that has not yet been fully explored.

The goal of this project is to broaden the knowledge of MMV networks and ultimately to deploy them in a real-world event-based processing use case, possibly combined with an event-based sensor.

[1] https://doi.org/10.1016/j.neunet.2018.08.014
[2] https://doi.org/10.1109/MSP.2019.2931595

Requirements

This project is intended for the inquisitive and creative engineering student, who is encouraged to follow his/her curiosity. Both theoretical and more hands-on approaches are encouraged.

Background in machine learning and digital design is required, as well as strong programming skills (e.g., Python and C).

This project will be in collaboration with Dr. Lars Keuninckx and Dr. Matthias Hartmann at IMEC Leuven and will be coupled with a 15-ECTS internship there. A relocation allowance during the internship will be offered.

Interested students should send a motivation letter together with their CV (incl. course transcripts and grades) to Dr. Charlotte Frenkel at c.frenkel@tudelft.nl

More MSc proposals for Dr. Charlotte Frenkel will appear in the coming weeks, interested students are encouraged to reach out by e-mail to enquire about upcoming projects.

Contact

Last modified: 2023-12-02