Papers
9
Total Citations
75
H-Index
5
About
Maximilian Gilles is a robotics and computer vision researcher whose work centers on autonomous robotic grasping, bin picking, and mobile robot perception. He is perhaps best known for pioneering the **MetaGraspNet** dataset series — large-scale, physics-based synthetic benchmarks designed to accelerate vision-driven robotic grasping research. The flagship MetaGraspNetV2 (2023, 27 citations) introduced a holistic framework enabling fast, reliable bin picking through object relationship reasoning and dexterous manipulation, while its predecessor (2022, 19 citations) laid the groundwork for scene-aware, ambidextrous grasping across diverse sensor modalities and gripper types. Beyond dataset construction, Gilles has made meaningful contributions to sim-to-real domain adaptation, developing active learning strategies and continual learning approaches that reduce the costly reliance on real-world training data. His more recent work, including vMF-Contact (2025), addresses uncertainty-aware grasp learning in cluttered, noisy environments — tackling both aleatoric and epistemic uncertainty for safer deployment. He has also contributed to structural SLAM and unsupervised ego-motion estimation, reflecting a broader interest in robust robot autonomy. With nearly 75 cumulative citations across nine publications, Gilles is an emerging voice shaping practical, deployment-ready robotic intelligence for smart manufacturing environments.
Research Focus
Key Achievements
Top Papers
- 1
- 2
- 3Structure SLAM with points, planes and objects7 citations · 2022
- 4
- 5
- 6
- 7
- 8
- 9