Yunchu Zhang

About

Yunchu Zhang’s research spans two transformative eras in robotics: from the rugged autonomy of power transmission line inspection to the frontier of data-driven robot manipulation. His early work (2006–2007) tackled the hard constraints of high-voltage environments, pioneering vision-based obstacle recognition, motion deblurring, and visual navigation for inspection robots—contributions that remain foundational, with papers like “Structure-Constrained Obstacles Recognition” and “Motion Based Image Deblur” each garnering over 100 cumulative citations. These projects demanded robust perception under kinematic uncertainty, which he addressed through neural network-based tracking control and sonar feature mapping. More recently, Zhang has become a key figure in large-scale robot learning. As a co-author of “DROID: A Large-Scale In-The-Wild Robot Manipulation Dataset” (2024, 108 citations), he helped create a benchmark that is reshaping how robots learn from diverse, real-world interactions. His work on “Energy-based Models are Zero-Shot Planners for Compositional Scene Rearrangement” (2023, 16 citations) introduces a paradigm where robots interpret compositional language instructions without task-specific training. Meanwhile, “Cherry-Picking with Reinforcement Learning” (2023, 11 citations) advances fine manipulation in cluttered, deformable environments—a challenge critical to surgery, harvesting, and assisted feeding. Zhang’s trajectory—from structured outdoor inspection to open-ended manipulation—exemplifies a rare versatility, bridging classical control and modern learning to expand what robots can achieve in the wild.

Research Focus

Key Achievements

Top Papers

1. 1
   DROID: A Large-Scale In-The-Wild Robot Manipulation Dataset

   108 citations · 2024
2. 2
   Structure-Constrained Obstacles Recognition for Power Transmission Line Inspection Robot

   27 citations · 2006
3. 3
   Energy-based Models are Zero-Shot Planners for Compositional Scene Rearrangement

   16 citations · 2023
4. 4
   Cherry-Picking with Reinforcement Learning

   11 citations · 2023
5. 5
   Motion Based Image Deblur Using Recurrent Neural Network for Power Transmission Line Inspection Robot

   11 citations · 2006
6. 6
   Visual Navigation for a Power Transmission Line Inspection Robot

   9 citations · 2006
7. 7
   Visual Navigation for a Power Transmission Line Inspection Robot

   8 citations · 2006
8. 8
   Tracking Control of a Mobile Robot with Kinematic Uncertainty Using Neural Networks

   6 citations · 2006
9. 9
   Sonar Feature Map Building for a Mobile Robot

   5 citations · 2007
10. 10
    Motion Deblurring for a Power Transmission Line Inspection Robot

    5 citations · 2006

Key Collaborators

Contact & Links