When: 23rd of April 2026, 1:00pm AEST
Where: This seminar will be partially presented at the ACFR seminar area, J04 lvl 2 (Rose St Building) and partially online via Zoom. RSVP
Speaker: Dr. Yiduo Wang
Title: Moving Towards Online Dynamic SLAM
Abstract:
Dynamic SLAM methods jointly estimate for the static and dynamic scene components. However existing approaches, while accurate, are computationally expensive and unsuitable for online applications, while the inherent unpredictability of dynamic entities demand online motion estimation. We therefore present the first application of incremental optimisation techniques to Dynamic SLAM, using a novel factor-graph formulation and system architecture designed to take advantage of existing incremental optimisation methods and support online estimation. On multiple datasets, we demonstrate that our method achieves equal to or better than state-of-the-art in camera pose and object motion accuracy, and show that our formulation results in problem structure well-suited to incremental solvers, while our system architecture further enhances performance, achieving a 5× speed-up over existing methods. Lately, we are exploring new directions to further decouple backend optimisation from a real-time front-end while maintaining accurate object motion estimation online.
Bio:
Dr. Yiduo Wang is a postdoctoral researcher in the ACFR, working with Dr Viorela Ila. He focuses on SLAM, reconstruction and navigation in dynamic environments. He holds a DPhil Engineering Science degree awarded by University of Oxford, where his research focused on large-scale reconstruction. He also holds an MRes Robotics degree awarded by UCL on combining SLAM with semantic segmentation for dynamic environments.