Yifan Liu

Yifan Liu

Ph.D. Candidate
Chinese University of Hong Kong liuyifan365[at]gmail.com

Research Interests:

  • Video/3D World Model
  • 3D Reconstruction
  • Generative AI

About Me

I am currently a final year Ph.D. student in the Department of Electronic Engineering, Chinese University of Hong Kong (CUHK), supervised by Prof. Yixuan Yuan. Before that, I received B. Eng. degree in Electronic Information Engineering from the Huazhong University of Science and Technology (HUST) in 2021. During my Ph.D. studies, I was fortunate to work with Dr. Zhenwei Wang and Dr. Tengfei Wang from Tencent Hunyuan, 3D Center, World Model Team.

News

  • [04/2026] Our paper WorldMirror is accepted by ICML 2026!
  • [10/2025] Our work WorldMirror is released. Check it out!
  • [02/2025] Our paper MonoSplat is accepted by CVPR'25.
  • [02/2025] Our paper EndoGaussian is accepted by TMI'25.
  • [01/2025] Our paper InstantSplamp is accepted by ICLR'25. Congrats to Chenxin and Hengyu.
  • [01/2025] Our paper Hide-in-Motion is accepted by ICRA'25. Congrats to Hengyu.
  • [12/2024] Our paper U-KAN is accepted by AAAI'25. Congrats to Chenxin.
  • [06/2024] Our LGS, Endora, and EndoSparse are accepted by MICCAI'24. Congrats to Hengyu and Chenxin.
  • [05/2024] Our paper SAMTooth is accepted by MICCAI'24 (early accept, 11%).
  • [03/2023] Our paper GRAB-Net is accepted by IEEE TMI'23. (IF. 11.037).

Experience

Selected Publications | Full list

(*equal contribution)

ICML 2026

WorldMirror: Universal 3D World Reconstruction with Any-Prior Prompting

Yifan Liu*, Zhiyuan Min*, Zhenwei Wang*, Junta Wu, Tengfei Wang, Yixuan Yuan, Yawei Luo, Chunchao Guo

International Conference on Machine Learning (ICML 2026)

Paper Code Project Page Try it now

A universal 3D reconstruction framework capable of processing diverse input modalities and generating various output representations.

CVPR 2025

MonoSplat: Generalizable 3D Gaussian Splatting from Monocular Depth Foundation Models

Yifan Liu, Keyu Fan, Weihao Yu, Chenxin Li, Hao LU, Yixuan Yuan

IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR 2025)

Paper Code

A generalizable 3D Gaussian Splatting approach from monocular depth foundation models.

ICLR 2025
InstantSplamp Paper

InstantSplamp: Fast and Generalizable Stenography Framework for Generative Gaussian Splatting

Chenxin Li*, Hengyu Liu*, Zhiwen Fan, Wuyang Li, Yifan Liu, Panwang Pan, Yixuan Yuan

International Conference on Learning Representations (ICLR 2025)

Paper Code Project Page

An initial exploration into embedding customizable, imperceptible, and recoverable information within the renders produced by off-the-line 3D generative models, while ensuring minimal impact on the rendered content's quality.

IEEE TMI 2025
Foundation Model-guided Gaussian Splatting Paper

Foundation Model-guided Gaussian Splatting for 4D Reconstruction of Deformable Tissues

Yifan Liu, Chenxin Li, Hengyu Liu, Chen Yang, Yixuan Yuan

IEEE Transactions on Medical Imaging (IEEE TMI), 2025

Paper Code Project Page

The first work to apply 3D Gaussian Splatting for dynamic (4D) endoscopic reconstruction, guided by foundation models.

AAAI 2025
U-KAN Paper

U-KAN Makes Strong Backbone for Medical Image Segmentation and Generation

Chenxin Li, Xinyu Liu, Wuyang Li, Cheng Wang, Hengyu Liu, Yifan Liu, Zhen Chen, Yixuan Yuan

AAAI Conference on Artificial Intelligence (AAAI 2025)

Paper Code Project Page

Introducing Kolmogorov-Arnold Networks (KANs) as a strong backbone for medical image segmentation and generation.

ICCV 2025
GaussianReg Paper

GaussianReg: Rapid 2D/3D Registration for Emergency Surgery via Explicit 3D Modeling with Gaussian Primitives

Weihao Yu, Xiaoqing Guo, Xinyu Liu, Yifan Liu, Haonan Zheng, Yawen Huang, Yixuan Yuan

IEEE/CVF International Conference on Computer Vision (ICCV 2025)

Paper Code

Rapid 2D/3D registration for emergency surgery using sparse Gaussian primitives and a ray-based registration scheme.

ECCV 2024
GTP-4o Paper

GTP-4o: Modality-prompted Heterogeneous Graph Learning for Omni-modal Biomedical Representation

Chenxin Li, Xinyu Liu, Cheng Wang, Yifan Liu, Weihao Yu, Jing Shao, Yixuan Yuan

European Conference on Computer Vision (ECCV 2024)

Paper Project Page

A heterogeneous graph framework for unified omni-modal biomedical representation that handles missing modalities.

ICCV 2023
FedPD Paper

FedPD: Federated Open Set Recognition with Parameter Disentanglement

Chen Yang, Meilu Zhu, Yifan Liu, Yixuan Yuan

IEEE/CVF International Conference on Computer Vision (ICCV 2023)

Paper

A federated open-set recognition framework with parameter disentanglement for cross-client closed/open-set learning.

Professional Activities

  • Conference Reviewer: NeurIPS, ICLR, ICML, CVPR, ICCV, ECCV, AAAI
  • Journal Reviewer: TPAMI, TIP, TVCG, TMM, TCSVT, TMI, MIA

Selected Awards

  • Self-Reliance Scholarship
    2019-2020
  • Scholarship of Academic Excellence
    2017-2020
  • Outstanding Graduates of Huazhong University of Science and Technology
    2021