Hello, I’m Chaoda Zheng, currently pursuing my Ph.D. at the Chinese University of Hong Kong, Shenzhen (CUHKSZ). I am privileged to conduct my research under the expert supervision of Dr. Zhen Li, and additionally co-supervised by the highly accomplished Prof. Shuguang Cui. My academic journey is guided by my passion for 3D computer vision, with a specific focus on perception tasks related to Autonomous Driving and Indoor Scene Understanding. However, I don’t limit myself to these areas; I am continuously exploring and broadening my knowledge in other related aspects such as neural rendering and multi-modal learning.

News

[2023/10]M2-Track extension is accepted by TPAMI.

[2023/08] LATR is selected as ICCV 2023 ORAL.

[2023/07] One paper on 3D point cloud upsampling is accepted by ACM MM 2023.

[2023/07] One paper on 3D lane detections is accepted by ICCV 2023.

[2023/06] Won 1st place in 4D Semantic Segmentation and 3rd place in 4D Action Segmentation at the CVPR 2023’s HOI4D Challenge.

[2023/01] CAT is accepted by TNNLS.

[2022/11] PointCMT is accepted by NeuraIPS 2022 (Spotlight). It can improve arbitrary point-based encoder via cross-modal training.

[2022/10] 2DPASS is accepted by ECCV 2022. With cross-modal distillation, our pure point-based network beats multi-modal networks in KITTI and NuScenes.

[2022/06] M2-Track is accepted by CVPR 2022 (Oral). The proposed motion-centric paradigm is Faster, Superior, and Memory-Efficient.

[2021/10] Won 2nd award in ICCV 2021 Challenge on Urban Scenes Understanding.

[2021/10] BAT is accepted by ICCV 2021. It enhances existing tracker with a simply-yet-effective box-aware feature enhancement.

[2021/6] CAR-NET is accepted by TIP. It ranks 1st on ModelNet40/10 and ShapeNetCore55 benchmarks.

[2020/6] Our robust and versatile point cloud operator PointASNL is accepted by CVPR 2020.

Activities

Reviewer

CVPR, ICCV, TIP

Talks

• Invited talk at TechBeat on Motion-Centric Single Object Tracking, June 2022. [Recording]

Teaching Assistant

• CUHKSZ CIE6032: Selected Topics in Deep Learning Foundations and Their Applications
• CUHKSZ CSC1001: Introduction to Computer Science: Programming Methodology
• CUHKSZ CSC4140: Computer Graphics