Yuheng Liu
About
I am a first-year CS Ph.D. student at Texas A&M University, supervised by Prof. Wenping Wang and Prof. Xin Li. Now, I am working closely with Prof. Xinke Li at City University of Hong Kong. I received my B.Eng. from Southwest Jiaotong University (CN) and B.Sc. from University of Leeds (UK) with First-class Degree, Outstanding Graduates of Sichuan Province and Best Bachelor Thesis in 2024. Furthermore, Iām a research intern at University of California, Merced, working with Prof. Lu Qi and Dr. Xueting Li, under the supervision of Prof. Ming-Hsuan Yang. Previously, I worked with Prof. Chongshou Li in SWJTU.
My research interests mainly lie in 3D vision and 3D generation. I am now focusing on 3D outdoor scene generation tasks, aiming to enhance the quality of generated outdoor scenes, make the generation process more controllable, and contribute to bridging the gap between synthetic environments and the complexities of the real world.
News
2024
November
Our work 'Controllable 3D Outdoor Scene Generation' is submitted. Access to our Project Page for more details.
2024
September
š¤ I received the Best Student award in my final year.
2024
August
Code of 'Pyramid Diffusion for Fine 3D Large Scene Generation' has been released. Access to our Github for more information.
2024
August
š Our work 'Pyramid Diffusion for Fine 3D Large Scene Generation' is accepted as an Oral Presentation!
2024
July
š Our work 'Pyramid Diffusion for Fine 3D Large Scene Generation' is accepted by ECCV 2024!
2024
June
š„° I graduated from SWJTU with Outstanding Graduates of Sichuan Province and Best Bachelor Thesis, and University of Leeds with First-class Degree!
2024
February
Our work 'Deep Hierarchical Learning for 3D Semantic Segmentation' is submitted. Access to our Project Page for more details.
2023
November
Our work 'Pyramid Diffusion for Fine 3D Large Scene Generation' is on arXiv.
2023
November
Our work 'Pyramid Diffusion for Fine 3D Large Scene Generation' is submitted. Access to our Project Page for more details.
2023
October
Awarded a national scholarship (top 0.2% nationwide).
2023
September
Ranked first in my major during my junior year.
Research & Publications
Pyramid Diffusion for Fine 3D Large Scene Generation
Yuheng Liu, Xinke Li, Xueting Li, Lu Qi, Chongshou Li, Ming-Hsuan Yang
ECCV 2024 (Oral Presentation, top 2%)
This work uses a coarse-to-fine approach with scale-varied diffusion models to generate high-quality outdoor scenes. It successfully generates 3D scenes both unconditionally and conditionally, and it allows for easy fine-tuning across different datasets and generating infinite scenes.
Controllable 3D Outdoor Scene Generation
Yuheng Liu, Xinke Li, Yuning Zhang, Xin Li, Lu Qi, Wenping Wang, Chongshou Li, Xueting Li, Ming-Hsuan Yang
Under Review, 2024
This work introduces a novel method that leverages scene graphs as an intuitive and user-friendly control format, enabling the generation of high-quality 3D outdoor scenes closely aligned with user-defined descriptions through a conditional diffusion framework.
Deep Hierarchical Learning for 3D Semantic Segmentation
Chongshou Li, Yuheng Liu, Xinke Li, Yuning Zhang, Tianrui Li, Junsong Yuan
Under Review, 2024
This work introduces Deep Hierarchical Learning (DHL) for 3D data, enhancing hierarchical embedding learning and class hierarchy construction for effective 3D object and scene segmentation.
Experience
University of California, Merced
Jun 2023 - present Ā· Research Assistant
Research Area: Diffusion Model, Out-door Scene Generation
Supervisor: Prof. Ming-Hsuan Yang
University of Leeds
Sep 2020 - Jun 2024 Ā· Undergraduate
B.Sc. in Computer Science
Graduated with Honors First-class Degree
Final Year: Best Student in Computer Science
Southwest Jiaotong University
Sep 2020 - Jun 2024 Ā· Undergraduate
B.Eng. in Computer Science
GPA: 91.02 / 100 (ranked 1 / 73 in year-2 and year-3)
Graduated with Outstanding Graduates of Sichuan Province & Best Bachelor Thesis
Honors & Awards
First-class Degree in Leeds, UK, 2024
Best Student in Computer Science, University of Leeds, 2024
Provincial Outstanding Graduate (top 3%), Sichuan Province, 2024
Outstanding Bachelor Thesis (top 1%, 1 student in major), SWJTU, 2024
National Scholarship (top 0.2% nationwide), Ministry of Education of the People's Republic of China, 2023
Pacemaker to Merit Student Award (top 1%), SWJTU, 2023
Lan Qiao Cup, Provincial First Prize, Sichuan Province, 2023
Collage Special First Prize Scholarship (top 1%, 1 student in major), SWJTU, 2022
Merit Student Award (top 5%), SWJTU, 2022
Chinese Computer Design Competition (Team Leader), National Second Prize, China, 2022
Comprehensive Scholarship, Third Prize, SWJTU, 2021
Talk
[10/03/2024] ECCV 2024 Oral Presentation: Pyramid Diffusion for Fine 3D Large Scene Generation, Milan, Italy.