Learning the Next Best View for 3D Point Clouds via Topological Features

Collander, Christopher; Beksi, William J.; Huber, Manfred

doi:10.1109/icra48506.2021.9561389

Cited by 9 publications

(1 citation statement)

References 26 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…As the robot is executing and re-planning a trajectory, determining where to aim the head camera-i.e. Next Best View planning for receding-horizon trajectory optimisation-is an interesting problem in itself which is more commonly explored in the context of building 3D models of objects or structures [34], [35], rather than for use in motion planning. Naïve heuristics, such as always looking in the direction of motion or looking a specified distance along the planned trajectory, are prone to failure cases, in particular on curved trajectories.…”

Section: Discussion and Future Workmentioning

confidence: 99%

Simultaneous Scene Reconstruction and Whole-Body Motion Planning for Safe Operation in Dynamic Environments

Finean,

Merkt,

Havoutis

2021

Preprint

View full text Add to dashboard Cite

Recent work has demonstrated real-time mapping and reconstruction from dense perception, while motion planning based on distance fields has been shown to achieve fast, collision-free motion synthesis with good convergence properties. However, demonstration of a fully integrated system that can safely re-plan in unknown environments, in the presence of static and dynamic obstacles, has remained an open challenge. In this work, we first study the impact that signed and unsigned distance fields have on optimisation convergence, and the resultant error cost in trajectory optimisation problems in 2D path planning, arm manipulator motion planning, and whole-body loco-manipulation planning. We further analyse the performance of three state-of-the-art approaches to generating distance fields (Voxblox, Fiesta, and GPU-Voxels) for use in realtime environment reconstruction. Finally, we use our findings to construct a practical hybrid mapping and motion planning system which uses GPU-Voxels and GPMP2 to perform recedinghorizon whole-body motion planning that can smoothly avoid moving obstacles in 3D space using live sensor data. Our results are validated in simulation and on a real-world Toyota Human Support Robot (HSR).

show abstract

Section: Discussion and Future Workmentioning

confidence: 99%