Deepak Rajamohan

Papers

Total Citations

H-Index

About

Deepak Rajamohan is a researcher at the forefront of computer vision, specializing in image-based localization and 3D scene understanding. His work addresses a critical challenge in robotics and augmented reality: achieving accurate pose estimation when there is a significant perspective difference between offline Structure from Motion (SfM) models and online SLAM keyframes. His most-cited paper, "Image based Localization under large perspective difference between SfM and SLAM using split sim(3) optimization" (2022, 6 citations), introduces a novel split sim(3) optimization framework that robustly aligns these disparate viewpoints, enabling reliable localization even when conventional methods fail. This contribution is particularly valuable for long-term autonomous navigation in dynamic environments. Rajamohan’s research bridges the gap between large-scale mapping and real-time tracking, with implications for self-driving cars, drone navigation, and AR applications. By tackling the co-visibility problem head-on, he has provided a practical solution that enhances the resilience of visual localization systems. His work continues to influence the development of more adaptive and robust perception pipelines in computer vision and robotics.