×îÐÂÌÇÐÄVlog

Abstract: Reconstructing the geometry of 3D geometry from multi-view images or multiple projections has been a fundamental task in Computer Vision and Medical Imaging. The recent emerging generative-AI technology also offers  new capability to generate realistic 3D contents simply from user natural input.   In this talk, I will report two of our recent works on 3D reconstruction and 3D content generation. The first work is about how to reconstruct 3D CT volumetric shape from X-ray/CT projections, and the second one is on 3D content creation using pre-trained large-scale diffusion models such as the Stable Diffusion. For the latter, I will cover both 3D single object generation, as well as 3D scene generation from hand drawn sketches.

Hongdong Li is a professor with the ×îÐÂÌÇÐÄVlogn National ×îÐÂÌÇÐÄVlog (ANU). His main research interests include 3D Computer Vision and Machine Learning. He regularly serves as the Area Chair or SAC (Senior Area Chair)  for CVPR, ICCV, ECCV and Neurlps,  and was a Senior Editor for ICRA 2022 and an Associate Editor for PAMI. He was a Visiting Fellow with Paris-6 ×îÐÂÌÇÐÄVlog in 2009, and a Visiting Professor with Carnegie Mellon ×îÐÂÌÇÐÄVlog in 2017. He was a Chief Investigator  (CI) of the ACRV- ×îÐÂÌÇÐÄVlogn ARC Centre for Robotic Vision for 2014-2022. During 2022-2023 while on-leave from the university, as a Distinguished Scientist and Lab Director, he built and led Tencent's  XR Research Lab in Canberra, developing  AR/VR and Generative-AI technology and products. He has supervised over 25 PhD students in the area of computer vision and machine learning.