Papers
14
Total Citations
1,694
H-Index
12
About
Kuan‐Ting Yu is a roboticist whose work has fundamentally advanced the ability of robots to perceive and manipulate objects in cluttered, real-world environments. His research sits at the intersection of computer vision, robotic grasping, and state estimation, with a particular focus on enabling robots to handle both known and novel objects without task-specific training data. Yu’s most influential contributions include his work on the Amazon Picking Challenge, where he developed a multi-view self-supervised deep learning system for 6D pose estimation that achieved 487 citations. He also pioneered a robotic pick-and-place system capable of multi-affordance grasping and cross-domain image matching, a breakthrough that has garnered over 460 citations. Beyond grasping, Yu has made significant contributions to robotic manipulation through high-fidelity datasets—such as his planar pushing dataset with over a million data points (163 citations)—and real-time state estimation frameworks that fuse tactile and visual sensing. His work on whole-body planning for the DARPA Robotics Challenge further demonstrates his ability to tackle complex, integrated robotic systems. With over 1,600 total citations across his top papers, Yu’s research continues to shape the future of autonomous manipulation in warehouses, homes, and beyond.
Research Focus
Key Achievements
Top Papers
- 1
- 2
- 3
- 4
- 5An Architecture for Online Affordance‐based Perception and Whole‐body Planning140 citations · 2014
- 6
- 7A Summary of Team MIT's Approach to the Amazon Picking Challenge 201543 citations · 2016
- 8
- 9
- 10