Danial Kamran

Karlsruhe Institute of Technology, Sharif University of Technology

Papers

Total Citations

H-Index

About

Danial Kamran is a robotics researcher whose work bridges the critical gap between simulation and real-world autonomy. His primary research areas include reinforcement learning for mobile robotics, vision-based navigation, and efficient motion estimation for autonomous systems. Kamran’s most impactful contribution is his pioneering work on sim-to-real transfer for path tracking, where he developed a Reinforcement Learning algorithm trained entirely in simulation and deployed it without modification on a real car-like robot—a 2019 paper that has garnered 22 citations and demonstrated a practical pathway for cost-effective robot training. He has also advanced vision-based navigation with his work on online visual gyroscopes for autonomous cars (7 citations), addressing the critical problem of rotation estimation when GPS signals are unavailable. Additionally, Kamran’s 2016 paper on 3-Point RANSAC using GPU technology (3 citations) introduced a computationally efficient method for fast rotation estimation by decoupling it from translation, improving sensor fusion algorithms. His work is notable for its focus on practical, deployable solutions that reduce reliance on expensive hardware or perfect environmental conditions, making autonomous navigation more accessible and robust.