About Me
I’m a 3nd-year M.Phil. student in Data Science and Information Technology at TBSI, Tsinghua University(THU), supervised by Prof. Lu Fang and Prof. Qionghai Dai. Prior to that, I spent wonderful 4 years at Sun Yat-Sen University(SYSU), and obtained my bachelor’s degree in 2021.
My research interests mainly lie in 3D computer vision, especially in:
- 3D Representation: Geometric Primitive based 3D Modeling (such as Quadrics and CAD)
- 3D Reconstruction: Semantic-aware Reconstruction and Rendering from images and video
- 3D Understanding: Hierarchical 3D Scene Parsing and Navigation
My goal is to model, reconstruct, and manipuate the 3D world in a more efficient and interpretable manner.
News
- [Feb. 2024] Our paper OmniSeg3D is accepted by CVPR 2024 and code is available now!
- [July. 2023] Our paper PARF is accepted by ICCV 2023!
- [July. 2022] Our paper ParseMVS is accepted by ACM MM 2022!
Publications
OmniSeg3D: Omniversal 3D Segmentation via Hierarchical Contrastive Learning Haiyang Ying, Yixuan Yin, Jinzhi Zhang, Fan Wang, Tao Yu, Ruqi Huang, Fu Fang†(CVPR 2024) | |
PARF: Primitive-Aware Radiance Fusion for Indoor Scene Novel View Synthesis Haiyang Ying, Baowei Jiang, Jinzhi Zhang, Di Xu, Tao Yu†, Qionghai Dai, Fu Fang†The International Conference on Computer Vision (ICCV) 2023 | |
ParseMVS: Learning Primitive-aware Surface Representations for Sparse Multi-view Stereopsis Haiyang Ying*, Jinzhi Zhang*, Yuzhe Chen, Zheng Cao, Jing Xiao, Ruqi Huang†, Fu Fang†The 30th ACM International Conference on Multimedia (ACM MM 2022) |
Research Experiences
Semantic-Aware Indoor Scene Reconstruction and Rendering Sep. 2022 – May. 2023 Propose a semantic-aware hybrid representation for indoor scene modeling.
Design a framework for fast indoor scene reconstruction and rendering with RGB-D input.
Implement the proposed framework based on Instant-NGP to achieve higher-quality rendering.
Mentor: Prof. Lu Fang and Tao YuSemantic-Aware Sparse View 3D Reconstruction Sep. 2021 – Aug. 2022
Propose a semantic-based representation to encode geometry, texture, and visibility on primitives.
Design a pipeline for multi-view 3D reconstruction under sparse observations.
Explore the capacity of implicit function for local representation and optimization.
Mentor: Prof. Lu Fang and Prof. Ruqi HuangDynamic Vascular 3D Reconstruction (B.S. Thesis) Oct. 2020 – Aug. 2021
Propose to model 3D dynamic tissue with implicit representation.
Design a pipeline for 3D reconstruction with multi-view cone beam CT images.
Mentor: Prof. Lu Fang
Education
- M.Phil., Data Science and Information Technology
Sept. 2021 - Present, Tsinghua University, China. - B.Eng., Electronic Information Science and Technology
Sept. 2017 - Jun. 2021, Sun Yat-Sen University, China.
Honors & Awards
- 2021, Excellent Graduate, Sun Yat-Sen University.
- 2019, 2020, National Scholarship, the Minister of Education, China.
- 2018, 1nd Prize, College Electronic Design Competition, Guangdong, China.
- 2018, 2019, 2020, Excellence Student the 1st-class Scholarship, Sun Yat-Sen University.
- 2018, AEON Scholarship, Sun Yat-Sen University.
Teaching Experiences
Fall 2022, Teaching Assistant, Tsinghua University.
Course: Media and Cognition (English)Fall 2023, Teaching Assistant, Tsinghua University.
Course: Machine Vision (English)