Papers
3
Total Citations
106
H-Index
3
About
Dahua Lin is a leading figure in computer vision and embodied AI, whose research bridges 3D perception, scene understanding, and human-robot interaction. His major contributions center on enabling machines to perceive and navigate complex 3D environments from limited sensory input. Notably, his work on monocular 3D object detection—using depth from motion—has been highly influential, with his 2022 paper accumulating 49 citations for tackling the fundamental challenge of predicting absolute depth from a single image. More recently, Lin introduced "EmbodiedScan" (2024, 54 citations), a holistic multi-modal 3D perception suite designed to empower embodied agents with the ability to explore environments, understand first-person observations, and contextualize scenes into language for interaction. This work represents a significant leap toward practical embodied AI. With a career marked by pioneering advances in 3D vision and a growing impact on robotics, Lin’s research continues to shape how machines see, understand, and act in the physical world.
Research Focus
Key Achievements
Top Papers
- 1EmbodiedScan: A Holistic Multi-Modal 3D Perception Suite Towards Embodied AI54 citations · 2024
- 2Monocular 3D Object Detection with Depth from Motion49 citations · 2022
- 3Monocular 3D Object Detection with Depth from Motion3 citations · 2022