ROCVJan 16, 2020

Probabilistic 3D Multilabel Real-time Mapping for Multi-object Manipulation

arXiv:2001.05752v13 citations
AI Analysis

This work addresses the need for efficient and accurate 3D mapping in robotics for handling multiple objects in cluttered settings, representing an incremental advancement over previous single-label methods.

The paper tackles the problem of real-time 3D multilabel object mapping for multi-object manipulation, achieving a 40-96% relative improvement over conventional methods in segmentation accuracy and enabling successful robot recognition (86.9%) and manipulation (60.7%) in occluded environments.

Probabilistic 3D map has been applied to object segmentation with multiple camera viewpoints, however, conventional methods lack of real-time efficiency and functionality of multilabel object mapping. In this paper, we propose a method to generate three-dimensional map with multilabel occupancy in real-time. Extending our previous work in which only target label occupancy is mapped, we achieve multilabel object segmentation in a single looking around action. We evaluate our method by testing segmentation accuracy with 39 different objects, and applying it to a manipulation task of multiple objects in the experiments. Our mapping-based method outperforms the conventional projection-based method by 40 - 96\% relative (12.6 mean $IU_{3d}$), and robot successfully recognizes (86.9\%) and manipulates multiple objects (60.7\%) in an environment with heavy occlusions.

Foundations

The foundational work for this paper's niche, ranked by how specifically the neighbourhood builds on it — not by global fame.

Your Notes