Ian Lenz
Papers
9
Total Citations
2,250
H-Index
7
About
Ian Lenz is a robotics and machine learning researcher whose work sits at the intersection of deep learning and robotic perception and control. He is best known for pioneering the application of deep learning to robotic grasp detection, a problem central to enabling robots to interact meaningfully with physical objects. His landmark 2015 paper, "Deep Learning for Detecting Robotic Grasps," has accumulated over 1,600 citations, establishing it as a foundational reference in robotic manipulation and demonstrating that learned feature representations could dramatically outperform hand-engineered alternatives for grasp prediction from RGB-D data. His work on DeepMPC extended these ideas into model predictive control, showing how deep latent features could tame complex nonlinear dynamics — including challenging tasks like robotic food-cutting — earning over 340 citations. Lenz also contributed to hierarchical semantic labeling for task-relevant robot perception and explored multimodal embedding frameworks that allow robots to reason jointly over point clouds, natural language, and motion trajectories. His doctoral thesis, "Deep Learning for Robotics," synthesizes these threads into a cohesive vision for data-driven robotic systems. Across his career, Lenz has helped lay the empirical and methodological groundwork for modern learned robotic manipulation.
Research Focus
Key Achievements
Top Papers
- 1Deep learning for detecting robotic grasps1,646 citations · 2015
- 2DeepMPC: Learning Deep Latent Features for Model Predictive Control344 citations · 2015
- 3Deep Learning for Detecting Robotic Grasps125 citations · 2013
- 4Hierarchical Semantic Labeling for Task-Relevant RGB-D Perception73 citations · 2014
- 5
- 6
- 7Deep Learning For Robotics11 citations · 2016
- 8
- 9Deep Learning for Detecting Robotic Grasps6 citations · 2013