Postdoctoral Researcher on Deep Learning at the Donders Centre for Cognition

Updated: about 1 month ago
Deadline: 09 May 2021

The department of AI at the Donders Institute is recruiting two postdoctoral researchers on a new project called SPIKEFERENCE, co-funded by the Human Brain Project. They will work under the supervision of Pablo Lanillos, Serge Thill and Marcel van Gerven. 

The SPIKEFERENCE project aims to control industrial robots with spiking neural networks using neuroscience-inspired algorithms (deep active inference), providing adaptation and conditional hierarchical problem-solving. The outcome of the project will be a step forward in brain-inspired computational models, both theoretically and experimentally, bridging the gap between low-level control and neurosymbolic approaches, contributing to major challenges within the Human-Brain Project, such as the development of neuroscience-driven adaptive cognitive architectures, and providing novel spike-driven closed-loop controllers that can be deployed in real robots.

This postdoctoral researcher position (deep learning profile) is framed in a new promising line of research within neuroscience-inspired artificial intelligence linked to adaptation and improvisation. Starting from our current deep active inference models, you will develop novel algorithms for robot perception and action by combining ideas from state representation learning, variational inference and functional models of the brain.. Initial prototyping and validation will take place in simulation (ROS-Gazebo and Neurorobotics platform). You will work in close collaboration with your supervisors, other postdoctoral researchers, and a PhD candidate.

Furthermore, you will join an exciting and vibrant young team of experts in machine learning, artificial intelligence and neurocomputing. As we support the Human Brain Project's equal opportunities policy, we strongly encourage submissions from various representative minorities.


View or Apply

Similar Positions