Dr Moritz Milde

International Centre for Neuromorphic Systems


Postdoctoral Research Fellow

Roles / Position Responsibilities

Software engineering, algorithm development, event-based computation, event-based processing


Dr Milde received his B.Sc. degree in biomimetics at the Westphalian University of Applied Science Bocholt in 2013. This line of studies focused on mechanical engineering with specification in light weight construction, as well as biology with specialization in sensory systems and bio-inspired information processing. He deepened his knowledge in biological sensory information processing during his M.Sc. in Neurobiology at Bielefeld University in 2015. Moritz worked together with Martin Egelhaaf and Elisabetta Chicca on the implementation of the visual information processing system of flying insects onto neuromorphic hardware and applied the abstracted artificial information processing network in the context of mobile robots in order to avoid collisions with surrounding objects.

Moritz completed his PhD in event-based computation and algorithm development at the Institute of Neuroinformatics at University of Zurich and ETH Zurich together with Giacomo Indiveri and Matthew Cook in 2019.

Research Interests

Dr Milde is interested into transferring insights of biological information processing onto state-of-the-art soft- and hardware. The transfer is not a one-to-one copy of the biological model, but rather an abstraction of the underlying principle in order to implement the developed neural circuits in Spiking Neural Networks (SNN).

Currently he is exploring and developing SNN architectures, which directly operate on local spatio-temporal contrast changes (events) rather than frames. These networks inherently represent time in their nature of computation. He is investigating how we can introduce a conceptual understanding of spatio-temporal patterns in the context of event-based sensory-processing systems.

The process of learning these spatio-temporal features is completely unsupervised and event-driven. The key insight, which also differentiates his approach from conventional machine learning, is that the network is continuously learning using STDP and that the error is expressed as failed prediction of the future in strictly local neighborhood.


  • PhD (Event-based computation & algorithm development), Institute of Neuroinformatics, University of Zurich & ETH Zurich. 2019
  • M.Science (Neurobiology) Bielefeld University, Germany. 2015
  • B.Science (Biomimetics), Westphalian University of Applied Sciences, Germany. 2013


A full listing of my publications can be found on my Google Scholar page here(opens in a new window)


LocationPenrith (Werrington South campus)
RoomBuilding BA.2.10