Muhammad Burhan Hafez
Researcher in Machine Learning and Robotics
UHH - Google Scholar - ResearchGate - GitHub
I am currently a Postdoctoral Research Associate in the Knowledge Technology Group at Universität Hamburg, working with Prof. Stefan Wermter. My research lies at the intersection of machine learning, robotics, and motor neuroscience, with a focus on robot motor learning. I investigate the role of the biological principles of self-organization and intrinsic motivation and the neurocognitive mechanisms of meta-decision making, strategy selection, and adaptive integration of model-based and model-free control in enabling robots to efficiently and continually acquire new skills. My research interests include:
Neural Networks
Reinforcement Learning
Cognitive Robotics
Announcements & News:
May 2023 - Our paper "Map-based Experience Replay: A Memory-Efficient Solution to Catastrophic Forgetting in Reinforcement Learning" has been accepted for publication in Frontiers in Neurorobotics. Preprint available here.
Apr 2023 - New preprint out: Model Predictive Control with Self-supervised Representation Learning (link).
Mar 2023 - New preprint out: Chat with the Environment: Interactive Multimodal Perception using Large Language Models (link).
Feb 2023 - I gave an invited talk titled "Efficient, Hybrid, and Continual Learning: Towards Brain-Inspired Intelligent Robots" at the Department of Computer Science, University of Bath, UK.
Feb 2023 - One paper on transformer-based multimodal learning accepted in Applied Artificial Intelligence (link).
Nov 2022 - I gave an invited talk titled "Towards Cognitive-Inspired Robot Skill Learning" at the Bernoulli Institute, University of Groningen, The Netherlands.
Jul 2022 - New paper on cross-modal self-supervised robot learning accepted at IROS 2022.
Nov 2021 - Invited talk on "Robotic Platform for Social Communication – Continual Robot Learning" at the Transregional Collaborative Research Centre on Cross-modal Learning, Hamburg, Germany.
Jul 2021 - My paper titled "Behavior Self-Organization Supports Task Inference for Continual Robot Learning" is accepted at IROS 2021.
Apr 2021 - Our paper titled "Improving Model-Based Reinforcement Learning with Internal State Representations through Self-Supervision" is accepted at IJCNN 2021.
Nov 2020 - My article titled "Improving robot dual-system motor learning with intrinsically motivated meta-control and latent-space experience imagination" has been published in Robotics and Autonomous Systems.
May 2020 - I successfully defended my PhD thesis "Intrinsically Motivated Actor-Critic for Robot Motor Learning" with magna cum laude.
Dec 2019 - Invited talk on "Robot Motor Learning with Intrinsically Motivated Actor-Critic" at TU-Delft, The Netherlands.