Object Grasping
Object grasping research aims to enable robots to reliably and efficiently manipulate objects, mirroring human dexterity. Current efforts focus on improving grasp planning and execution through various methods, including deep learning models for grasp detection and pose prediction from visual and tactile data, reinforcement learning for policy optimization, and the development of novel gripper designs. These advancements are crucial for applications in robotics, assistive technologies, and virtual reality, promising to enhance automation in diverse fields and improve human-robot interaction.
Papers
Sim-to-Real Grasp Detection with Global-to-Local RGB-D Adaptation
Haoxiang Ma, Ran Qin, Modi shi, Boyang Gao, Di Huang
ALDM-Grasping: Diffusion-aided Zero-Shot Sim-to-Real Transfer for Robot Grasping
Yiwei Li, Zihao Wu, Huaqin Zhao, Tianze Yang, Zhengliang Liu, Peng Shu, Jin Sun, Ramviyas Parasuraman, Tianming Liu