Francesco Verdoja
University of Turin, Aalto University, University of Technology
Papers
9
Total Citations
103
H-Index
4
About
Francesco Verdoja is a roboticist whose research sits at the intersection of 3D perception, semantic mapping, and dexterous manipulation. His work addresses fundamental challenges in enabling autonomous robots to understand and interact with complex, unstructured environments. A central contribution is his pioneering work on 3D point cloud segmentation, where his 2017 paper on fast supervoxel-based segmentation—garnering 45 citations—established an efficient method for scene understanding using geometry and color from RGB-D cameras. He has also made significant strides in robotic mapping, authoring a comprehensive 2023 survey on Maps of Dynamics (27 citations) that synthesizes how robots can represent and reason about changing environments. In manipulation, Verdoja has advanced the field of multi-fingered grasping, introducing generative sampling methods like DDGC and Multi-FinGAN to enable dexterous grasping in clutter. His work on uncertainty-aware navigation, using deep networks to overcome the limitations of 2D laser scanners, further demonstrates his commitment to robust, real-world robotic systems. With a growing portfolio of highly cited papers and a focus on creating smarter, updatable map representations, Verdoja is shaping the future of autonomous robotics.
Research Focus
Key Achievements
Top Papers
- 1
- 2Survey of maps of dynamics for mobile robots27 citations · 2023
- 3Deep Network Uncertainty Maps for Indoor Navigation14 citations · 2019
- 4Constrained Generative Sampling of 6-DoF Grasps5 citations · 2023
- 5DDGC: Generative Deep Dexterous Grasping in Clutter3 citations · 2021
- 6Online Object-Oriented Semantic Mapping and Map Updating3 citations · 2021
- 7Multi-FinGAN: Generative Coarse-To-Fine Sampling of Multi-Finger Grasps3 citations · 2021
- 8On the Potential of Smarter Multi-layer Maps2 citations · 2020
- 9