Jingnan Gao 高景南

I am a third-year Ph.D. student at AI Institute, Shanghai Jiao Tong University (SJTU), supervised by Prof. Yichao Yan and Prof. Xiaokang Yang. I received my bachelor's degree at Shanghai Jiao Tong University (SJTU) in 2023.

My current interest lies in 3D computer vision and graphics, especially in 3D geometry reconstruction/generation.

Email  /  Google Scholar  /  Github

profile photo

News

  • News (Feb 2026): Two papers are accepted to CVPR 2026.
  • News (Jan 2026): Two papers are accepted to ICLR 2026.
  • News (Jan 2025): One paper is accepted to ICLR 2025.
  • News (Nov 2024): One paper is accepted to TVCG 2024.
  • News (Sep 2024): One paper is accepted to NeurIPS 2024.
  • News (Sep 2024): One paper is accepted to Computer Graphics Forum (CGF) 2024.

Experiences
alibaba
2025 - 2026
Research Intern
Alibaba Meta Team
  • Host by Jiangjing Lyu, working closely with Xianze Fang and Zhe Wang.
  • Topic: 3D Foundation Model and Mesh Generation.
    • tencent
    2024 - 2025
    Research Intern
    Tencent XR Vision Labs
  • Host by Weizhe Liu and Pan Ji.
  • Topic: Mesh Detailization and Mesh Generation.
    • alibaba
    2023 - 2024
    Research Intern
    Alibaba Meta Team
  • Host by Jiangjing Lyu.
  • Topic: Efficient 3D Reconstruction.

    • Selected Publications
    clean-usnob MeshPack: Towards High-Resolution Artistic Mesh Generation via Part-wise Autoregressive Modeling
    Qinglin Yang, Xianze Fang, Jingnan Gao, Jiangjing Lyu, Yichao Yan, Xiaowei Zhou, Jiazhi Xia, Sida Peng
    project page

    MeshPack efficiently generates part-level high-resolution artistic meshes conditioned on point clouds.

    clean-usnob MoRE: 3D Visual Geometry Reconstruction Meets Mixture-of-Experts
    Jingnan Gao*, Zhe Wang*, Xianze Fang, Xingyu Ren, Zhuo Chen, Shengqi Liu, Yuhao Cheng, Jiangjing Lyu, Xiaokang Yang, Yichao Yan
    CVPR, 2026
    Integrated into the Taobao3D (Alibaba) project for 3D geometry priors estimation.
    project page / arXiv / code

    MoRE is a feed-forward foundation model that leverages mixture-of-experts in 3D visual geometry reconstruction. MoRE takes unposed images as input and achieving robust geometric predictions for various scenarios.

    clean-usnob POLAR: A Portrait OLAT Dataset for Relightable Face Modeling
    Zhuo Chen*, Chengqun Yang*, Zhuo Su, Zheng Lv, Jingnan Gao, Xiaoyuan Zhang, Xiaokang Yang, Yichao Yan
    CVPR, 2026
    project page / arXiv / dataset / code

    POLAR is a large-scale and physically calibrated One-Light-at-a-Time (OLAT) dataset containing over 200 subjects captured under 156 lighting directions, multiple views, and diverse expressions.

    clean-usnob Dens3R: A Foundation Model for 3D Geometry Prediction
    Xianze Fang*, Jingnan Gao*, Zhe Wang, Zhuo Chen, Xingyu Ren, Jiangjing Lyu, Qiaomu Ren, Zhonglei Yang, Xiaokang Yang, Yichao Yan, Chengfei Lyu
    ICLR, 2026
    Contributed to the Taobao3D (Alibaba) project as a 3D foundation model.
    project page / arXiv / code

    Dens3R is a feed-forward visual foundation model that takes unposed images as input and outputs high-quality 3D pointmap with unified geometric dense prediction.

    clean-usnob SurfSplat: Conquering Feedforward 2D Gaussian Splatting with Surface Continuity Priors
    Bing He, Jingnan Gao, Yunuo Chen, Ning Cao, Gang Chen, Zhengxue Cheng, Li Song, Wenjun Zhang
    ICLR, 2026
    project page / arXiv / code

    SurfSplat reconstructs coherent and realistic 3D surfaces using 2DGS, enabled by a surface continuity prior and forced alpha blending.

    clean-usnob Fused-Granularity Neural Surfaces with Anisotropic Encoding for High-Fidelity 3D Reconstruction
    Zhuo Chen*, Jingnan Gao*, Xiaokang Yang, Yichao Yan
    Submitted to IJCV, 2026
    Extension of AniSDF (ICLR 2025).

    AniSurf reconstructs high-fidelity 3D geometry and significantly improves the rendering quality of both SDF-based method and UDF-based method.

    clean-usnob Relightable and Animatable Gaussian Head Avatar from Monocular Videos
    Zhuo Chen, Yichao Yan, Jingnan Gao, Zhuo Su, Chao Wen, Zhaohu Li, Yuhao Cheng, Xueying Lee, Yutong Leng, Yikun Zeng, Guidong Wang, Xiaokang Yang
    Submitted to TVCG, 2026

    Our method can reconstruct a photorealistic avatar together with its high-quality materials from monocular video only.

    clean-usnob MARS: Mesh AutoRegressive Model for 3D Shape Detailization
    Jingnan Gao, Weizhe Liu, Weixuan Sun, Senbo Wang, Xibin Song, Taizhang Shang, Shenzhou Chen, Hongdong Li, Xiaokang Yang, Yichao Yan, Pan Ji
    Submitted to CVMJ, 2026
    arXiv

    MARS generates intricate geometric patterns while preserving the overall shape integrity of the coarse mesh input.

    clean-usnob AniSDF: Fused-Granularity Neural Surfaces with Anisotropic Encoding for High-Fidelity 3D Reconstruction
    Jingnan Gao, Zhuo Chen, Xiaokang Yang, Yichao Yan
    ICLR, 2025
    project page / arXiv / code

    AniSDF reconstructs high-fidelity 3D geometry and significantly improves the rendering quality of the SDF-based method.

    clean-usnob EvaSurf: Effcient View-Aware Implicit Textured Surface Reconstruction
    Jingnan Gao, Zhuo Chen, Yichao Yan, Bowen Pan, Zhe Wang, Jiangjing Lyu, Xiaokang Yang
    TVCG, 2024
    Deployed on the Taobao (Alibaba) platform.
    project page / arXiv

    EvaSurf reconstructs accurate 3D geometry and supports real-time rendering on mobile devices.

    clean-usnob Multi-times Monte Carlo Rendering for Inter-reflection Reconstruction
    Tengjie Zhu*, Zhuo Chen*, Jingnan Gao, Yichao Yan, Xiaokang Yang
    NeurIPS, 2024
    project page / arXiv / code

    Ref-MC2 employs multi-times monte carlo rendering for inter-reflection reconstruction.

    clean-usnob ITEM3D: Directional Texture Editing for 3D Models
    Shengqi Liu*, Zhuo Chen*, Jingnan Gao, Yichao Yan, Wenhan Zhu, Jiangjing Lyu, Xiaokang Yang
    Computer Graphics Forum (CGF), 2024
    Received Wiley China Excellent Author award (July-September, 2024).
    project page / pdf / code

    ITEM3D leverages Diffusion to achieve 3D texture editing.

    Selected Awards

    • Outstanding Graduate Student Scholarship, 2025.
    • Wiley China Excellent Author, 2024.
    • Outstanding Graduates, Shanghai Jiao Tong University, 2023.
    • Third Prize, ACM Multimedia 2022 Vico Conversational Head Generation Challenge, 2022.
    • Finalists Award, Mathematic Contest in Modeling(MCM), 2021.

    Services

    Miscellanea

    Thanks to Jon Barron for the template, and to Yinuo for taking my profile photo.

    Web Site Hit Counters
    Hit Counters