Papers
7
Total Citations
2,654
H-Index
6
About
Tanner Schmidt is a leading researcher in computer vision and robotics, specializing in 6D object pose estimation, articulated object tracking, and self-supervised learning for dense correspondence. His most impactful contribution is **PoseCNN**, a convolutional neural network that revolutionized 6D object pose estimation in cluttered, occluded scenes—a critical capability for robots interacting with the real world. This seminal work has garnered over **2,200 citations**, underscoring its foundational role in the field. Schmidt also developed **DART** (Dense Articulated Real-Time Tracking), a framework that enables real-time tracking of articulated objects (e.g., robot arms, human hands) using consumer depth cameras, with over **190 citations** across its publications. His research on self-supervised visual descriptor learning for dense correspondence (170 citations) advanced robust pixel-level matching for tracking and mapping. By integrating physical constraints into depth-based tracking, Schmidt enhanced robot manipulation in realistic scenarios. His work bridges perception and action, providing tools for robots to understand and interact with dynamic, unstructured environments—a cornerstone for autonomous systems.
Research Focus
Key Achievements
Top Papers
- 1
- 2Self-Supervised Visual Descriptor Learning for Dense Correspondence170 citations · 2016
- 3DART: Dense Articulated Real-Time Tracking136 citations · 2014
- 4
- 5Depth-based tracking with physical constraints for robot manipulation69 citations · 2015
- 6DART: dense articulated real-time tracking with consumer depth cameras57 citations · 2015
- 7Self-directed Lifelong Learning for Robot Vision3 citations · 2019