Papers
39
Total Citations
1,230
H-Index
14
About
Karl Van Wyk is a robotics researcher whose work sits at the intersection of dexterous manipulation, robot learning, and human-robot interaction. He is best known for his foundational contributions to benchmarking and datasets in robotic grasping — most notably the DexYCB dataset (2021, 250 citations), which became a standard reference for evaluating hand-object interaction across 2D detection, 6D pose estimation, and keypoint localization. His DexPilot system (2020, 197 citations) demonstrated that vision-based teleoperation of multi-fingered robotic hands could be achieved affordably without sacrificing dexterity, opening new pathways for imitation learning and skill transfer. Van Wyk has also advanced sim-to-real transfer through DeXtreme (2023) and contributed to scalable motion planning with CuRobo (2023), a GPU-parallelized collision-free trajectory generation framework. His early work on force-based manipulation and assembly benchmarking protocols helped establish rigorous, reproducible evaluation standards for fine manipulation tasks — a theme that runs throughout his career. With contributions spanning geometric control theory, in-hand pose tracking, and teleoperation systems, Van Wyk has built a cohesive research identity around making dexterous robotic manipulation both measurable and deployable in real-world settings.
Research Focus
Key Achievements
Top Papers
- 1DexYCB: A Benchmark for Capturing Hand Grasping of Objects250 citations · 2021
- 2DexPilot: Vision-Based Teleoperation of Dexterous Robotic Hand-Arm System197 citations · 2020
- 3
- 4DeXtreme: Transfer of Agile In-hand Manipulation from Simulation to Reality88 citations · 2023
- 5
- 6Benchmarking Protocols for Evaluating Small Parts Robotic Assembly Systems82 citations · 2020
- 7CuRobo: Parallelized Collision-Free Robot Motion Generation82 citations · 2023
- 8
- 9
- 10