Dinkar Gupta

Papers

Total Citations

H-Index

About

Dinkar Gupta is a researcher whose work lies at the intersection of computer vision and robotics, with a particular focus on structure-from-motion and ego-motion estimation. His most-cited paper, "Planar Ego-Motion Without Correspondences" (2005, 6 citations), addresses a critical limitation in traditional vision-based localization: while general structure-from-motion methods struggle with constrained camera motions—such as those found in mobile robotics—existing ego-motion techniques rely heavily on establishing feature correspondences between images. Gupta’s contribution introduces a novel approach that bypasses this requirement, enabling more efficient and robust motion estimation for planar camera trajectories. This work is particularly valuable for applications in autonomous navigation and robot localization, where computational efficiency and reliability are paramount. Though his citation count is modest, the conceptual innovation of eliminating correspondence-based constraints marks a meaningful step forward in simplifying vision tasks for constrained environments. Gupta’s research demonstrates a keen understanding of practical challenges in real-world robotic systems, offering solutions that bridge the gap between theoretical computer vision and applied robotics.