Yunxin Tai
Papers
2
Total Citations
8
H-Index
2
About
Yunxin Tai is a rising researcher in robotics and artificial intelligence, with a primary focus on visual learning for robotic manipulation. Their work centers on developing end-to-end solutions that enable robots to learn complex grasping and manipulation tasks directly from visual data, reducing the need for extensive manual programming. Tai's most notable contribution is the "Grasp Proposal Networks," a pioneering approach that integrates visual learning with robotic grasp planning, achieving significant progress in 6-degree-of-freedom grasping for parallel-jaw grippers. This work, published in 2020, has garnered 5 citations and laid the groundwork for more efficient robotic learning systems. More recently, Tai introduced "You Only Teach Once," a method for one-shot bimanual robotic manipulation learning from video demonstrations, published in 2025 with 3 citations. This innovative approach demonstrates how robots can acquire complex bimanual skills from a single video example, pushing the boundaries of sample efficiency in robot learning. Tai's research is particularly impactful for students and researchers interested in bridging computer vision and robotics, offering practical frameworks that reduce data requirements while maintaining robust performance in real-world manipulation tasks.
Research Focus
Key Achievements
Top Papers
- 1
- 2