Stefano Gasperini


Foto von Stefano Gasperini

Stefano Gasperini

PhD Candidate in 3D Scene Understanding for Autonomous Driving

Postadresse

Postal:
Boltzmannstr. 3
85748 Garching b. München

 LinkedIn          Google Scholar

Research Interests

  • Autonomous Driving
  • 3D Computer Vision
  • Reliable Scene Understanding
  • Robustness
  • Domain Generalization
  • 3D Reconstruction

Publications

  • Alexander Lehner; Stefano Gasperini; Alvaro Marcos-Ramiro; Michael Schmidt; Nassir Navab; Benjamin Busam; Federico Tombari: 3D Adversarial Augmentations for Robust Out-of-Domain Predictions. International Journal of Computer Vision (IJCV), 2023 mehr…
  • Felix Tristram; Stefano Gasperini; Federico Tombari; Nassir Navab; Benjamin Busam: Re-Nerfing: Enforcing Geometric Constraints on Neural Radiance Fields through Novel Views Synthesis. arXiv preprint, under review, 2023 mehr…
  • Stefano Gasperini; Nils Morbitzer; HyunJun Jung; Nassir Navab; Federico Tombari: Robust Monocular Depth Estimation under Challenging Conditions. IEEE/CVF International Conference on Computer Vision (ICCV), 2023 mehr…
  • Sen Wang; Wei Zhang; Stefano Gasperini; Shun-Cheng Wu; Nassir Navab: VoxNeRF: Bridging Voxel Representation and Neural Radiance Fields for Enhanced Indoor View Synthesis. arXiv preprint, under review, 2023 mehr…
  • Stefano Gasperini; Alvaro Marcos-Ramiro; Michael Schmidt; Nassir Navab; Benjamin Busam; Federico Tombari: Segmenting Known Objects and Unseen Unknowns without Prior Knowledge. IEEE/CVF International Conference on Computer Vision (ICCV), 2023 mehr…
  • Alexander Lehner; Stefano Gasperini; Alvaro Marcos-Ramiro; Michael Schmidt; Mohammad-Ali Nikouei Mahani; Nassir Navab; Benjamin Busam; Federico Tombari: 3D-VField: Adversarial Augmentation of Point Clouds for Domain Generalization in 3D Object Detection. IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 2022 mehr…
  • Stefano Gasperini; Jan Haug; Mohammad-Ali Nikouei Mahani; Alvaro Marcos-Ramiro; Nassir Navab; Benjamin Busam; Federico Tombari: CertainNet: Sampling-free Uncertainty Estimation for Object Detection. IEEE Robotics and Automation Letters (RA-L), 2021 mehr…
  • Stefano Gasperini; Mohammad-Ali Nikouei Mahani; Alvaro Marcos-Ramiro; Nassir Navab; Federico Tombari: Panoster: End-to-end Panoptic Segmentation of LiDAR Point Clouds. IEEE Robotics and Automation Letters (RA-L), 2021 mehr…
  • Stefano Gasperini; Patrick Koch; Vinzenz Dallabetta; Nassir Navab; Benjamin Busam; Federico Tombari: R4Dyn: Exploring Radar for Self-Supervised Monocular Depth Estimation of Dynamic Scenes. Proceedings of the IEEE International Conference on 3D Vision (3DV), 2021, 2021 mehr…
  • Stefano Gasperini; Magdalini Paschali; Carsten Hopke; David Wittmann; Nassir Navab: Signal Clustering with Class-independent Segmentation. Proceedings of the IEEE International Conference on Acoustics, Speech, and Signal Processing (ICASSP), 2020 mehr…
  • Magdalini Paschali; Stefano Gasperini; Abhijit Guha Roy; Michael Y.-S. Fang; Nassir Navab: 3DQ: Compact Quantized Neural Networks for Volumetric Whole Brain Segmentation. Proceedings of the International Conference on Medical Image Computing and Computer Assisted Intervention (MICCAI), 2019 mehr…

Curriculum Vitae

  • 11/2019 - ongoing: PhD Candidate in Reliable 3D Scene Understanding for Autonomous Driving at TUM,
    supervised by PD Dr. Ing. Habil. Federico Tombari. PhD thesis submitted in December 2023.
  • 11/2019 - 12/2022: PhD Researcher in Autonomous Driving at BMW, Unterschleißheim, Germany.
  • 04/2019 - 10/2019: Master Thesis Student at Airbus, Manching, Germany.
  • 04/2018 - 03/2019: Working Student Researcher at Airbus, Manching, Germany.
  • 10/2017 - 11/2019: Master in Informatics at TUM.
  • 10/2016 - 09/2017: Software Developer at Prometeia / Modis, Bologna, Italy.
  • 09/2013 - 10/2016: Bachelor in Computer Engineering at the University of Bologna, Italy.
  • 09/2015 - 06/2016: Exchange Student at the University of California Santa Cruz (UCSC), California, USA.

Scholarships and Awards

  • Outstanding Reviewer at ICCV 2023 (top 1.87% of 6990 reviewers).
  • Deutschlandstipendium holder, merit-based Scholarship program (top 1% of students in Germany), for the second consecutive year, sponsored by the German government and Airbus (10.2018-09.2019).
  • Deutschlandstipendium holder, merit-based Scholarship program (top 1% of students in Germany), sponsored by the German government and a TUM association (10.2017-09.2018).
  • Graduated “summa cum Laude” ("110 e Lode") in Computer Engineering, University of Bologna, Italy (07.10.2016).
  • Provost’s Commendations for Excellence, University of Bologna, Italy (06.2015).

Student Projects

Supervised the following MSc student projects (end):

  • MSc Thesis - Neural Radiance Fields (NeRF) in highly sparse settings - Nils Morbitzer (ongoing).
  • Guided Research - Time-consistent generic object pose estimation - Tarlan Bakirli (2023).
  • MSc Thesis - Weakly Supervised Semantic Segmentation for RGB-D Data, Focus on Contrastive Learning and Pseudo Labeling - Mert Kiray - co-supervision with Lennart Bastian (2023).
  • MSc Thesis - Weakly Supervised Semantic Segmentation for RGB-D Data, Focus on Robustness and Oversegmentation - Merve Karali - co-supervision with Lennart Bastian (2023).
  • MSc Thesis - Self-supervised Segmentation for LiDAR Point Clouds - Philipp Nikutta - co-supervision with Lennart Bastian (2023).
  • IDP - Improving Robustness of Self-supervised Monocular Depth Estimation Methods - Nils Morbitzer (2023). 
  • MSc Thesis - Multi-modal Panoptic Segmentation for Autonomous Driving - Christopher Heumann - TUM and BMW sides (2022).
  • MSc Thesis - Open-Set Panoptic Segmentation for Autonomous Driving - Frithjof Winkelmann - TUM and BMW sides (2022).
  • MSc Thesis - Uncertainty Estimation for Object Detection in Autonomous Driving - Jan Haug - TUM and BMW sides (2021).
  • MSc Thesis - Enhancing Monocular Depth Estimation for Autonomous Driving with Sparse Radar Data - Patrick Koch - TUM and BMW sides (2021).
  • MSc Thesis - Evaluation of Semantic Segmentation Methods for LiDAR Point Clouds - Florian Hartmann - BMW side (2021).

Additionally supervised practical semester projects through the lab courses (Teaching Section). Topics include autonomous driving perception and 3D reconstruction.

Professional Services

Reviewed for the following conferences and journals:

  • CVPR: IEEE/CVF Conference on Computer Vision and Pattern Recognition
  • ICCV: IEEE/CVF International Conference on Computer Vision,
    Outstanding Reviewer at ICCV 2023 (top 1.87% of 6990 reviewers).
  • ECCV: Springer European Conference on Computer Vision
  • IJCV: Springer International Journal of Computer Vision
  • RA-L: IEEE Robotics and Automation Letters
  • ICRA: IEEE International Conference on Robotics and Automation
  • IROS: IEEE International Conference on Intelligent Robots and Systems
  • PR: Elsevir Pattern Recognition, journal
  • MedIA: Elsevir Medical Image Analysis Journal
  • MICCAI (indirectly): Springer International Conference on Medical Image Computing and Computer Assisted Intervention