3D Perception for Autonomous System
Abstract
Perception module enables the autonomous system with increased intelligence and the ability to adaptively collect and process sensor data into useful and actionable information, with the goal of minimizing or eliminating human intervention. Despite the continuous remarkable acceleration of deep learning solutions for computer vision, there are many challenges when applying deep learning to autonomous system in an open environment. In this talk, we will discuss current trends and challenges in 3D perception with a focus on point cloud scene understanding.
Brief Biography
Amani Alonazi, Artificial Intelligence Scientist of Boeing Research and Technology (BR&T). Amani is based in the Boeing office at King Abdullah University of Science and Technology (KAUST) —where she is an alumna. Her technical focus is on computer vision, 3D deep learning, machine learning, and HPC. Her research interests are to apply fundamental concepts of AI, optimization, and computer science to address important problems in aerospace, including UAV, robotic vision, and autonomous systems. She is a member of KAUST's Visual Computing Research Center, KAUST’s Robotics, Intelligent Systems, and Control lab, KAUST’s Extreme Computing Research Center, and ACM. Prior to joining Boeing, Amani received a Ph.D. degree from KAUST in 2019. Her dissertation focuses on asynchronous iterative algebraic solvers. Her work was recognized - 20K core hour award on Summit from Oak Ridge Leadership Computing, Best Research Award from ISC, and published in several top-rank conferences. She also has worked on large-scale graph processing, heterogeneous architectures, and video understanding and learning. In 2013, Amani held a one-year Erasmus scholarship from the European Union, which was spent at the School of Computer Science in University College Dublin, from which she received a second MS degree in Computer Science in 2014.