Jun Lv

Papers

Total Citations

H-Index

About

Jun Lv is a robotics and computer vision researcher whose work sits at the intersection of robot learning, 6D object pose estimation, and intelligent manipulation systems. His research addresses some of the most fundamental challenges in building capable, general-purpose robotic systems that can operate effectively in complex, real-world environments. Lv's most notable contribution is the SAGCI-System (2022), a framework designed to advance robot learning along four critical dimensions: sample efficiency, generalizability, compositionality, and incrementality. This ambitious work tackles the core bottlenecks preventing robots from achieving human-level performance across diverse tasks and environments, earning 21 citations and establishing Lv as a thoughtful voice in the pursuit of scalable robot intelligence. His earlier work, 6-PACK (2019), demonstrates his strong foundation in 3D perception. This deep learning system enables real-time, category-level 6D pose tracking on RGB-D data using learned anchor-based keypoints — allowing robots to track novel object instances across categories like bowls, laptops, and mugs without object-specific training. The paper has accumulated 12 citations and reflects Lv's ability to bridge perception and manipulation research. Together, these contributions position Jun Lv as an emerging researcher working toward robots that are not only perceptually aware but also adaptable and continuously learning.