SafePicking: Learning Safe Object Extraction via Object-Level Mapping
Not on the map. Why?
Abstract (arXiv)
Robots need object-level scene understanding to manipulate objects while reasoning about contact, support, and occlusion among objects. Given a pile of objects, object recognition and reconstruction can identify the boundary of object instances, giving important cues as to how the objects form and support the pile. In this work, we present a system, SafePicking, that integrates object-level mapping and learning-based motion planning to generate a motion that safely extracts occluded target objects from a pile. Planning is done by learning a deep Q-network that receives observations of predicted poses and a depth-based heightmap to output a motion trajectory, trained to maximize a safety metric reward. Our results show that the observation fusion of poses and depth-sensing gives both better performance and robustness to the model. We evaluate our methods using the YCB objects in both simulation and the real world, achieving safe object extraction from piles.
Key facts. Cited by 3 papers in the map (data through 7 Oct 2026). Builds on 26 of its 37 references that are in the map. First posted on arXiv 11 Feb 2022.
Builds on
26 of 37 references.
References · 26
- 2017Attention Is All You Needcited by 59,932
- 2014Adam: A Method for Stochastic Optimizationcited by 36,960
- 2019PyTorch: An Imperative Style, High-Performance Deep Learning Librarycited by 20,160
- 2017Mask R-CNNcited by 7,259
- 2015End-to-End Training of Deep Visuomotor Policiescited by 847
- 2017PoseCNN: A Convolutional Neural Network for 6D Object Pose Estimation in Cluttered Scenescited by 768
- 2018QT-Opt: Scalable Deep Reinforcement Learning for Vision-Based Robotic Manipulationcited by 730
- 2016Learning Hand-Eye Coordination for Robotic Grasping with Deep Learning and Large-Scale Data Collectioncited by 312
- 2019Emergent Tool Use From Multi-Agent Autocurriculacited by 308
- 2018Deep Object Pose Estimation for Semantic Robotic Grasping of Household Objectscited by 237
- 2015Supersizing Self-supervision: Learning to Grasp from 50K Tries and 700 Robot Hourscited by 208
- 2019kPAM: KeyPoint Affordances for Category-Level Robotic Manipulationcited by 152
14 more
- 2019TossingBot: Learning to Throw Arbitrary Objects with Residual Physicscited by 131
- 2017Robotic Pick-and-Place of Novel Objects in Clutter with Multi-Affordance Grasping and Cross-Domain Image Matchingcited by 124
- 2018Learning Synergies between Pushing and Grasping with Self-supervised Deep Reinforcement Learningcited by 121
- 2021Coarse-to-Fine Q-attention: Efficient Learning for Visual Robotic Manipulation via Discretisationcited by 120
- 2017Intel RealSense Stereoscopic Depth Camerascited by 82
- 2016Multi-view Self-supervised Deep Learning for 6D Pose Estimation in the Amazon Picking Challengecited by 70
- 2019Form2Fit: Learning Shape Priors for Generalizable Assembly from Disassemblycited by 51
- 2020MoreFusion: Multi-object Reasoning for 6D Pose Estimation from Volumetric Fusioncited by 46
- 2017Multi-Task Domain Adaptation for Deep Learning of Instance Grasping from Simulationcited by 35
- 2018Grasp2Vec: Learning Object Representations from Self-Supervised Graspingcited by 27
- 2019kPAM-SC: Generalizable Manipulation Planning using KeyPoint Affordance and Shape Completioncited by 27
- 2020Visuomotor Mechanical Search: Learning to Retrieve Target Objects in Cluttercited by 16
- 2019GRIP: Generative Robust Inference and Perception for Semantic Robot Manipulation in Adversarial Environmentscited by 6
- 2020GeoFusion: Geometric Consistency informed Scene Estimation in Dense Cluttercited by 5
Built on by
Citing papers per year: 2022: 3.
3 papers.