Grasping Object

Robotic grasping research aims to enable robots to reliably manipulate objects, mirroring the dexterity of human hands. Current efforts focus on improving grasp detection and planning using various approaches, including deep learning models (e.g., those incorporating 3D foundation models, VQDIF for shape completion, and transformer architectures for improved accuracy and robustness in noisy environments), and advanced control strategies leveraging tactile feedback and adaptive control methods. These advancements are crucial for expanding robotic capabilities in diverse applications, such as assistive robotics, warehouse automation, and aerial manipulation.

Papers