Mohit Singh

Norwegian University of Science and Technology

Papers

Total Citations

H-Index

About

Mohit Singh is a leading researcher in underwater robotics, specializing in state estimation and sensor fusion for autonomous navigation in challenging aquatic environments. His work bridges the gap between classical physics-based models and modern deep learning, with a core focus on visual-inertial odometry (VIO) and proprioceptive sensing. Singh’s major contributions include the development of **DeepVL**, a novel deep learning framework that predicts robot-centric velocity using dynamics-aware proprioception—leveraging inertial measurements, motor commands, and battery voltage through recurrent neural networks. This approach enables robust odometry without relying on visual features, a critical advantage in turbid or feature-poor underwater settings. He has also pioneered online self-calibrating refractive camera models, allowing VIO systems to estimate the refractive index of unknown media in real-time, thereby eliminating the need for pre-calibration or known targets. These contributions have garnered attention, with his most cited works accumulating over 10 citations since 2024. Singh’s achievements include enabling underwater robots to operate across diverse fluids with varying refractive properties, significantly advancing the reliability of long-duration subsea missions. His work is foundational for students and researchers aiming to push the boundaries of autonomous underwater exploration.