Marcel Büsching

I am a Ph.D. student at the Royal Institute of Technology (KTH) in the Division of Robotics, Perception and Learning (RPL) in Stockholm, under the supervision of Assoc. Prof. Mårten Björkman and co-supervision of Asst. Prof. Kathlén Kohn and Prof. Danica Kragic. My primary research interest is in the area of neural representation and rendering of dynamic scenes.

My Ph.D. position is funded by the grant 3D Scene Perception, Embeddings and Neural Rendering from the Wallenberg AI, Autonomous Systems and Software Program (WASP). For this reason, you can also meet me at courses of the WASP Graduate School.

news

Aug 31, 2023	Study Trip to the United States
Jul 01, 2023	Started my PhD in Stockholm
Mar 15, 2023	Finished my masters at Karlsruhe Institute of Technology

selected publications

FlowIBR: Leveraging Pre-Training for Efficient Neural Image-Based Rendering of Dynamic Scenes

Marcel Büsching, Josef Bengtson, David Nilsson, and 1 more author

arXiv preprint, arXiv.2309.05418, Sep 2023

Abs HTML

We introduce a novel approach for monocular novel view synthesis of dynamic scenes. Existing techniques already show impressive rendering quality but tend to focus on optimization within a single scene without leveraging prior knowledge. This limitation has been primarily attributed to the lack of datasets of dynamic scenes available for training and the diversity of scene dynamics. Our method FlowIBR circumvents these issues by integrating a neural image-based rendering method, pre-trained on a large corpus of widely available static scenes, with a per-scene optimized scene flow field. Utilizing this flow field, we bend the camera rays to counteract the scene dynamics, thereby presenting the dynamic scene as if it were static to the rendering network. The proposed method reduces per-scene optimization time by an order of magnitude, achieving comparable results to existing methods - all on a single consumer-grade GPU.
Distilling Semantic Features for 3D Cloth Representations from Vision Foundation Models

Alberta Longhini, Marcel Büsching, Bardienus Pieter Duisterhof, and 3 more authors

ICRA 2024 Workshop on 3D Visual Representations for Robot Manipulation, Sep 2024

Abs HTML

This study explores the use of vision foundation models to enhance 3D representations of cloth-like deformable objects. By focusing on the distillation of semantic information from RGB images, we examine the potential of pre-trained Visual-Language Models in capturing complex folded configurations of cloth. Our investigation reveals the challenges and preliminary successes in leveraging semantic information to improve the understanding and tracking of deformable object states.