Speaker

Dr. Hussein Bin Samma

Affiliate, Control and Instrumentation Engineering Department

Research Engineering III, SDAIA-KFUPM JRC for Artificial Intelligence– General, King Fahd University of Petroleum and Minerals

Dhahran, Saudi Arabia

Abstract

This presentation introduces the fundamentals of Reinforcement Learning (RL) and its extension to Deep Reinforcement Learning (DRL), focusing on key concepts such as the Deep Q-Learning (DQL) algorithm, Artificial Neural Networks (ANNs), and Convolutional Neural Networks (CNNs) for visual encoding. Building on these principles, two case studies are presented: training a robotic arm to perform a pick task, and enabling a drone to navigate using vision only. These studies demonstrate how DRL can learn complex behaviors directly from sensory data, highlighting its effectiveness and potential in advancing intelligent robotic systems.

Speaker Bio:

Dr. Hussein Samma is a Research Scientist at the SDAIA-KFUPM Joint Research Center for AI. He is affiliated with the Department of Control & Instrumentation Engineering. He holds a PhD. in Computer Engineering (Computer Vision, 2016), an MSc. in Computer Engineering (2009), and a BS. in Computer Engineering (2006). His research focuses on the intersection of deep reinforcement learning, computer vision, large language models, and their application for drone and robotics. At KFUPM, he has taught several AI-focused courses such as Data Science and AI for Robotics. At SDAIA-KFUPM JRC-AI, he leads multiple research projects, including "Autonomous Drone Visual Navigation using Deep RL" and "AI-Powered Crowd Analytics from Drone Streaming".