Dylan Campbell

Dylan Campbell is a Lecturer in Computing at the Australian National University. His research spans computer vision, optimisation, machine learning, and robotics, with a focus on 3D vision and deep learning optimisation. He has experience in geometric sensor data alignment, neural radiance fields, and differentiable optimisation layers, and is actively seeking motivated students for research projects.

Senior Lecturer ARC DECRA Fellow VComp Convener PhB Convener

Research interests: Computer vision, optimisation, machine learning, robotics, 3D vision, optimisation for deep learning, geometric sensor data alignment, camera localisation, SLAM, structure from motion, optical flow, 3D representations, neural radiance fields, differentiable optimisation layers, symmetries in data, efficient training of neural networks

Dylan Campbell is a Senior Lecturer at the Australian National University, specializing in 3D Vision, Generative Models, 3D Reconstruction, Optimisation, and Registration. His research focuses on developing globally optimal solutions for 3D point-set registration, such as Go-ICP and GOGMA, and explores methods for accurate camera geo-localization and robust point-set merging. He also investigates human-robot interaction through robotic vision and human-object interaction detection using transformer-based models. His work includes datasets like the IKEA ASM Dataset and techniques for spatially conditioned graph modeling. Campbell's research emphasizes geometric optimization, deep learning, and cross-view matching for applications in robotics, computer vision, and 3D scene reconstruction.