Marwan Shaker

About

Marwan Shaker is a researcher whose work lies at the intersection of autonomous robotics, computer vision, and reinforcement learning (RL). His primary research focus is on developing intelligent control systems that allow unmanned aerial vehicles (UAVs) and mobile robots to learn complex skills directly from visual input. Shaker’s most significant contribution is his pioneering approach to vision-based autonomous landing for UAVs, where he demonstrated that a simulated aircraft could learn to land from scratch by interacting with its environment using fast reinforcement learning. This work, published in 2010 with 21 citations, addresses one of the most challenging problems in UAV autonomy—precision landing without human intervention. He further advanced the field by exploring approximate policy iteration methods to accelerate the learning process for real-world robotic applications, a critical step toward making RL practical for physical systems. Shaker also developed AltURI, a thin middleware designed to optimize real-time vision processing in robot simulators, enabling faster and more efficient development of vision-based control applications. Through these contributions, Shaker has helped bridge the gap between simulated learning and real-world robotic performance, making him a notable figure in the ongoing effort to create truly autonomous aerial and ground robots.

Research Focus

Key Achievements

Top Papers

1. 1
   Vision-Based Landing of a Simulated Unmanned Aerial Vehicle with Fast Reinforcement Learning

   21 citations · 2010
2. 2
   Vision-based reinforcement learning using approximate policy iteration

   6 citations · 2009
3. 3
   AltURI: a thin middleware for simulated robot vision applications

   3 citations · 2011

Key Collaborators

Contact & Links