I am currently a Tech Lead Manager at Goertek Alpha Labs, where I lead research and development in XR (AR/VR/MR) and AI, with a particular focus on Spatial Content Generation and Spatial Intelligence technologies. My research interests lie in deep learning, with applications in 3D vision, robotic vision, and spatial AI.

Previously, I was a Staff Research Engineer/Scientist at InnoPeak Technology, Inc. (also known as OPPO US Research Center), working on advanced XR-related R&D projects. Prior to that, I served as a Postdoctoral Researcher at the University of Adelaide, collaborating with Prof. Ian Reid and Dr. Hamid Rezatofighi at Monash University. During this time, I was affiliated with both the Australian Institute for Machine Learning (AIML) and the Vision & Learning for Autonomous AI Lab (VL4AI).

I received my Ph.D. from the University of Adelaide, where I was part of the Australian Centre for Robotic Vision and was advised by Prof. Ian Reid. Earlier in my academic journey, I earned a B.Eng. in Electronic Engineering (First Class Honours) from The Chinese University of Hong Kong (CUHK), advised by Prof. Xiaogang Wang. I was also a visiting researcher in the Unmanned Systems Research Group at The National University of Singapore, working under Prof. Ben M. Chen.

NEWS

• 06/2026: One preprint is online: Understanding while Exploring: Semantics-driven Active Mapping [arxiv]

• 03/2026: One preprint is online: Semantic Consistent Language Gaussian Splatting for Point-Level Open-vocabulary Querying [arxiv]

• 02/2026: One paper is accepted to CVPR 2025: ActiveGAMER: Active GAussian Mapping through Efficient Rendering [project page] [arxiv] [code] [video]

• 01/2026: Start a new role at Goertek Alpha Labs.

• 05/2025: One paper is accepted to CGI 2025: Dynamic voxel grid optimization for high-fidelity rgb-d supervised surface reconstruction [arxiv]

• 01/2025: One paper is accepted to ICRA 2025: PlanarNeRF: Online Learning of Planar Primitives with Neural Radiance Fields [arxiv]

• 03/2024: One paper is accepted to CVPR 2024: NARUTO: Neural Active Reconstruction from Uncertain Target Observations. [project page] [paper] [arxiv] [code] [video]

• 12/2023: We are hiring 2024 research interns (US) and 2024 research interns (China) to work on research projects related to 3D Vision! Feel free to drop me an email if you’re interested.

• 10/2023: One paper is accepted to T-PAMI: SC-DepthV3: Robust Self-supervised Monocular Depth Estimation for Dynamic Scenes [paper] [arxiv] [code]

• 02/2023: We are hiring 2023 summer interns to work on research projects related to 3D vision! Feel free to drop me an email if you’re interested.

• 12/2022: It is my pleasure to be recognised as “Outstanding Reviewer” for ACCV 2022 for having provided helpful, high-quality reviews.

• 11/2022: Four preprints are online:

  SC-DepthV3: Robust Self-supervised Monocular Depth Estimation for Dynamic Scenes [arxiv]

  ActiveRMAP: Radiance Field for Active Mapping And Planning [arxiv]

  Predicting Topological Maps for Visual Navigation in Unexplored Environments [arxiv]

  What Images are More Memorable to Machines? [arxiv]

• 10/2022: Move to CA, USA and start my new position in InnoPeak Technology, Inc (a.k.a. OPPO US Research).

• 12/2021: One paper is accepted to 3DV: NVSS: High-quality Novel View Selfie Synthesis [paper]

• 12/2021: One paper is accepted to TPAMI: Auto-Rectify Network for Unsupervised Indoor Depth Estimation [paper] [arxiv] [code]

• 05/2021: One paper is accepted to IJCV: Unsupervised Scale-consistent Depth Learning from Video [paper] [arxiv] [code]

• 03/2021: The extended report for our ICRA2020 (DF-VO) is online: DF-VO: What Should Be Learnt for Visual Odometry? [arxiv] [code]

• 08/2020: Start my Postdoc position in The University of Adelaide.

• 06/2020: One preprint is online: Auto-Rectify Network for Unsupervised Indoor Depth Estimation [arxiv]

• 01/2020: One paper accepted to ICRA 2020: Visual odometry revisited: What should be learnt? [paper] [arxiv] [code] [video]

• 10/2019: One paper accepted to ICCV-Workshop (Deep Learning for Visual SLAM) 2019: Camera relocalization by exploiting multi-view constraints [paper]

• 09/2019: One paper accepted to NeurIPS 2019: Scale-consistent depth and ego-motion learning [paper] [arxiv] [code]

• 05/2019: Attend ICRA 2019 @ Montreal, Canada

• 01/2019: One paper accepted to ICRA 2019: Self-supervised depth and surface normal learning [paper] [arxiv]

• 07/2018: Join HoloLens team @ Microsoft Redmond as a Research Intern

• 07/2018: One paper accepted to ECCV 2018: Efficient dense point cloud object reconstruction [paper]

• 06/2018: Attend CVPR 2018 @ Salt Lake City, USA

• 02/2018: One paper accepted to CVPR 2018: Unsupervised monocular depth and visual odometry learning

• 06/2017: One paper accepted to IROS 2017: Deep learning for 2D scan matching and loop closure [paper]

• 02/2017: Start my Ph.D in The University of Adelaide