KCIS Invited Talk by Dr. Sourav Garg
on
Image Representation & Learning for Mobile Robots
Perception and navigation are fundamental capabilities of an autonomous mobile robot. Sensory data in the form of RGB camera images is information-rich, convenient to use, low-cost and power-efficient. With the advent of deep learning, processing raw images into meaningful representations has made a significant stride. However, depending on the end-task, for example, those relevant for mobile robot localization, such latent representations can be defined in many different ways to cover several unique aspects of autonomous navigation. This talk will cover my research in the field of visual place recognition, discussing the use of semantics, monocular depth, hashing, sequences, and deep learning to generate image representations that address various challenges in long-term robot autonomy.
Currently co-leading the Perception and Localization research theme at the QUT Centre for Robotics (QCR) as a Postdoctoral Research Fellow. He received his PhD from Queensland University of Technology (QUT) in 2019 while being a part of the Australian Centre for Robotic Vision (ACRV). Sourav pioneered research in the twin challenge of visual place recognition for mobile robot localization that requires dealing with scene appearance and camera viewpoint simultaneously. His research spans several aspects of robotic vision including semantics, sequences, hashing, adverse conditions, representation learning and centimeter-precise localization. His thesis received Executive Dean’s Commendation for Outstanding Thesis Award and his research published in Robotics: Science & Systems (RSS) and International Journal of Robotics Research (IJRR) won the SAGE QUT Student Publication Prize. He has published research at various top-tier international conferences and journals across both robotics and computer vision. He has also delivered invited talks at notable venues. Sourav has extensive academic research collaborations worldwide and has worked with both industry and government. His research has also been covered by numerous social media platforms including Brisbane Times, Engineers Australia Create Magazine, and Tech Xplore.
Venue: A3-117, Vindhya Block
Date and Time: 11 July, 3:00pm