5 papers
VoxAfford: Multi-Scale Voxel-Token Fusion for Open-Vocabulary 3D Affordance Detection
Haowen Sun, Shaolong Zhang, Mingyang Li +6
Open-vocabulary 3D affordance detection requires localizing interaction regions on point clouds given novel affordance descriptions. Recent methods extend multimodal large language…
Bridging Simulation and Reality: Cross-Domain Transfer with Semantic 2D Gaussian Splatting
Jian Tang, Pu Pang, Haowen Sun +4
Cross-domain transfer in robotic manipulation remains a longstanding challenge due to the significant domain gap between simulated and real-world environments. Existing methods suc…
PRISM: Projection-based Reward Integration for Scene-Aware Real-to-Sim-to-Real Transfer with Few Demonstrations
Haowen Sun, Han Wang, Chengzhong Ma +4
Learning from few demonstrations to develop policies robust to variations in robot initial positions and object poses is a problem of significant practical interest in robotics. Co…
DexDiff: Towards Extrinsic Dexterity Manipulation of Ungraspable Objects in Unrestricted Environments
Chengzhong Ma, Houxue Yang, Hanbo Zhang +5
Grasping large and flat objects (e.g. a book or a pan) is often regarded as an ungraspable task, which poses significant challenges due to the unreachable grasping poses. Previous…
REGNet V2: End-to-End REgion-based Grasp Detection Network for Grippers of Different Sizes in Point Clouds
Binglei Zhao, Han Wang, Jian Tang +5
Grasping has been a crucial but challenging problem in robotics for many years. One of the most important challenges is how to make grasping generalizable and robust to novel objec…