Zikang Shan
Papers
2
Total Citations
96
H-Index
2
About
Zikang Shan is a leading researcher in robotic dexterous manipulation, with a primary focus on universal grasping and goal-conditioned policy learning. His most impactful work, "UniDexGrasp," introduces a groundbreaking framework for learning universal robotic dexterous grasping directly from point cloud observations in table-top settings. This approach enables robots to generate diverse, high-quality grasps and lift objects across hundreds of categories, including unseen ones, by combining diverse proposal generation with a goal-conditioned policy. With 94 citations, this work has become a cornerstone in the field, demonstrating significant generalization capabilities beyond traditional grasping methods. Shan's contributions address critical challenges in robotic manipulation, such as adaptability to novel objects and environments, pushing the boundaries of what dexterous hands can achieve. His research is highly influential for students and researchers interested in reinforcement learning, computer vision, and robotics, offering a scalable solution for real-world applications like automated manufacturing and assistive robotics. Through UniDexGrasp, Shan has established himself as a key innovator in universal robotic grasping.
Research Focus
Key Achievements
Top Papers
- 1
- 2