Papers
26
Total Citations
597
H-Index
12
About
Timothy Patten is a robotics researcher whose work sits at the intersection of autonomous perception, multi-robot coordination, and robotic manipulation. He is perhaps best known for his foundational contributions to decentralized multi-robot active perception, most notably the Dec-MCTS algorithm, a decentralized variant of Monte Carlo Tree Search that enables teams of robots to collaboratively optimize their actions through shared probabilistic planning — a paper that has garnered over 186 citations and established him as a leading voice in the field. His broader research portfolio spans object detection, pose estimation, and 3D scene understanding, with significant contributions to template matching for texture-less objects, synthetic-to-real transfer learning, and active object classification using RGB-D data. Patten has also advanced robotic grasping through experience-based learning, introducing geometrical correspondence networks that allow robots to generalize from past successes to novel objects. His work extends into human-robot interaction, including intuitive VR-based teleoperation of humanoid robots and learning manipulation skills from human demonstrations. Collectively, his research addresses core challenges in enabling robots to perceive, reason, and act intelligently in complex, unstructured real-world environments.
Research Focus
Key Achievements
Top Papers
- 1Dec-MCTS: Decentralized planning for multi-robot active perception186 citations · 2018
- 2
- 3Monte Carlo planning for active object classification41 citations · 2017
- 4Viewpoint Evaluation for Online 3-D Active Object Classification40 citations · 2015
- 5
- 6Decentralised Monte Carlo Tree Search for Active Perception34 citations · 2020
- 7
- 8Hand-Object Interaction: From Human Demonstrations to Robot Manipulation24 citations · 2021
- 9
- 10