Marwan Taher

Hi! I am a 2nd year PhD student at the Dyson Robotics Lab supervised by Andrew J. Davison. I am intetretsted in scence represenation and understanding for downstream robotic manipulation.

In 2021, I completed my MEng in Mechatronics and Robotics at The University of Sheffiled, during which I worked on Long-term Outdoor Mapping for my master’s dissertation. During my time at university and previously during high school I have been very active in competitive robotics, participating in competitions such as MATE ROV and RoboCup.

Email  /  CV  /  Scholar  /  <Twitter  /  LinkdIn  /  Github

profile photo


I'm interested in robotics, computer vision and machine learning. Particularly the intersection between them to enable autnoumous manipulation via visuomotor control in the real-world.

Fit-NGP: Fitting Object Models to Neural Graphics Primitives
Marwan Taher, Ignacio Alzugaray, Andrew J. Davison
International Conference on Robotics and Automation (ICRA) , 2024.
project page / video / arXiv

Accurate 6 DoF object pose estimation via fitting known object models to Instant-NGP’s Density Field.

EscherNet: A Generative Model for Scalable View Synthesis
Xin Kong, Shikun Liu, Xiaoyang Lyu, Marwan Taher, Xiaojuan Qi , Andrew J. Davison

project page / arXiv

EscherNet is a multi-view conditioned diffusion model for view synthesis. EscherNet learns implicit and generative 3D representations coupled with the camera positional encoding (CaPE), allowing continuous relative camera control between an arbitrary number of reference and target views.

vMAP: Vectorised Object Mapping for Neural Field SLAM
Xin Kong, Shikun Liu, Marwan Taher, Andrew J. Davison
IEEE / CVF Computer Vision and Pattern Recognition Conference (CVPR) , 2023.
project page / video / arXiv

We present vMAP, an object-level real-time mapping system, with each object represented by a separate MLP neural field model, and object models are optimised in parallel via vectorised training.

Robust and long-term monocular teach and repeat navigation using a single-experience map
Li Sun, Marwan Taher, Christopher Wild, Cheng Zhao, Yu Zhang, Filip Majer, Zhi Yan, Tomáš Krajník, Tony Prescott, Tom Duckett
International Conference on Intelligent Robots and Systems (IROS) , 2021.

This paper presents a robust monocular visual teach-and-repeat (VT&R) navigation system for long-term operation in outdoor environments.

Website template courtesy of Jon Barron.