Johann Huber
Papers
4
Total Citations
27
H-Index
2
About
Johann Huber is a roboticist pushing the boundaries of data-driven robotic grasping, with a focus on overcoming the sparse reward and data bottleneck challenges that have long hindered the field. His research centers on the intersection of robotic manipulation, Quality-Diversity (QD) algorithms, and sim-to-real transfer. Huber’s major contribution is the innovative application of QD methods—traditionally used in evolutionary robotics for locomotion—to the domain of grasping. He has demonstrated that QD can generate diverse, high-performing grasp solutions even under conditions of sparse interaction and reward, a critical advance for real-world robotics. His work has produced key resources, including the QDGSet, a large-scale synthetic grasping dataset, and he has shown how domain randomization can effectively bridge the sim-to-real gap for automatically generated data. With his most cited paper, "Domain Randomization for Sim2real Transfer of Automatically Generated Grasping Datasets," already garnering 20 citations since 2024, Huber is establishing himself as a rising leader in making robotic grasping more robust, data-efficient, and generalizable.
Research Focus
Key Achievements
Top Papers
- 1
- 2Speeding up 6-DoF Grasp Sampling with Quality-Diversity3 citations · 2024
- 3
- 4Qdgset: a Large Scale Grasping Dataset Generated With Quality-Diversity2 citations · 2025