Jean-Baptiste Weibel

About

Jean-Baptiste Weibel is a computer vision and robotics researcher whose work centers on 3D object perception, 6D pose estimation, and deep learning for robotic systems. His research addresses some of the most pressing challenges in enabling robots to understand and interact with the physical world, particularly through monocular and RGB-based sensing pipelines. Weibel has made notable contributions to 6D object pose estimation, including the development of ZS6D, a zero-shot framework leveraging Vision Transformers to estimate object poses without object-specific training — a significant step toward generalizable robotic perception. His investigations into the challenges of monocular pose estimation have garnered over 30 citations, reflecting strong community interest. He has also tackled the notoriously difficult problem of transparent object perception, proposing ReFlow6D, which uses refraction-guided intermediate representations to estimate poses of otherwise nearly invisible objects. Beyond pose estimation, Weibel has contributed to 3D object classification through hybrid point pair feature and graph convolution methods, and explored the sim-to-real gap affecting learning-based approaches. His annotation toolkit, 3D-DAT, further supports the broader research community in building high-quality robotic vision datasets. With a growing citation record across multiple research threads, Weibel is establishing himself as a versatile and impactful contributor to robot perception research.

Research Focus

Key Achievements

Top Papers

1. 1
   Computer Vision Systems

   33 citations · 2023
2. 2
   Challenges for Monocular 6-D Object Pose Estimation in Robotics

   32 citations · 2024
3. 3
   ZS6D: Zero-shot 6D Object Pose Estimation using Vision Transformers

   32 citations · 2024
4. 4
   3D-DAT: 3D-Dataset Annotation Toolkit for Robotic Vision

   11 citations · 2023
5. 5
   Robust 3D Object Classification by Combining Point Pair Features and Graph Convolution

   8 citations · 2019
6. 6
   Challenges of Depth Estimation for Transparent Objects

   6 citations · 2023
7. 7
   Measuring the Sim2Real Gap in 3D Object Classification for Different 3D Data Representation

   5 citations · 2021
8. 8
   ReFlow6D: Refraction-Guided Transparent Object 6D Pose Estimation via Intermediate Representation Learning

   4 citations · 2024
9. 9
   Robust Sim2Real 3D Object Classification Using Graph Representations and a Deep Center Voting Scheme

   4 citations · 2022
10. 10
    Challenges for Monocular 6D Object Pose Estimation in Robotics

    2 citations · 2023

Key Collaborators

Contact & Links