Amir Ramezani Dooraki

About

Amir Ramezani Dooraki is a researcher advancing the frontier of autonomous robotics through deep reinforcement learning. His work centers on developing intelligent agents capable of navigating and exploring unknown, challenging environments without human intervention. Dooraki’s major contributions include pioneering end-to-end deep reinforcement learning controllers that enable robots to autonomously explore static and dynamic settings, such as environments with moving walls, nets, and pipes. His most cited paper, “An End-to-End Deep Reinforcement Learning-Based Intelligent Agent Capable of Autonomous Exploration in Unknown Environments” (36 citations), establishes a foundational framework for self-directed robotic exploration. He further extended this work with a multi-objective reinforcement learning controller (23 citations) that simultaneously optimizes multiple rewards for robust navigation. Dooraki also introduced memory-based reinforcement learning algorithms to enhance exploration efficiency and developed the bio-inspired FE-TRPO (Flight Enhanced Trust Region Policy Optimization) algorithm for multi-rotor flight. His research bridges artificial intelligence and robotics, pushing toward practical, resilient autonomous systems for applications in surveillance, rescue, and beyond.

Research Focus

Key Achievements

Top Papers

1. 1
   An End-to-End Deep Reinforcement Learning-Based Intelligent Agent Capable of Autonomous Exploration in Unknown Environments

   36 citations · 2018
2. 2
   A Multi-Objective Reinforcement Learning Based Controller for Autonomous Navigation in Challenging Environments

   23 citations · 2022
3. 3
   Memory-based reinforcement learning algorithm for autonomous exploration in unknown environment

   10 citations · 2018
4. 4
   Multi-rotor Robot Learning to Fly in a Bio-inspired Way Using Reinforcement Learning

   5 citations · 2019

Key Collaborators

Contact & Links