Homepage · Delong Wu (武德龙)

About & Background

Education

Nanjing University 985 Double First-Class

Sep 2023 – Jun 2026

M.Eng. in Electronic Information (Communication Engineering) · School of Electronic Science and Engineering

GPA 90.89/100 (Rank 2/48) National Scholarship 2025 First-Class Academic Scholarship (×2)

Hohai University 211

Sep 2019 – Jun 2023

B.Sc. in Communication Engineering · School of Information Science and Engineering

GPA 4.8/5.0 (Rank 5/128) Academic Excellence Scholarship (×4 consecutive years)

Industry Experience

Alibaba Taobao Tmall Group Research Intern

May – Aug 2025

Business Technology Department · 3D Reconstruction & Generative Algorithms

Core Contributor → Home3D 1.0

Research Interests

3D Reconstruction & Generation Recovering and synthesizing high-fidelity 3D geometry, PBR materials, and structured parts from images, video, or point clouds.
Implicit Neural Representations Encoding scene properties such as SDF geometry, radiance, wavelet primitives, and motion as continuous neural fields.
Computer Vision Enabling machines to interpret visual signals, spanning multi-view geometry, neural rendering, and physical signal blending.
Artificial Intelligence Diffusion transformers, flow matching, and deep optimization methods for scalable visual and spatial intelligence.

Selected Publications

* indicates equal contribution / alphabetical order

Home3D 1.0: A High-Fidelity Image-to-3D Asset Generation System for Interior Design Tech Report 2026

Yiyun Fei*, Guoqiu Li*, Jin Song*, Chuqiao Wu*, Delong Wu*, Hong Wu*, Ziru Zeng*, Haohui Chen, Yindong Kong, Jing Li, Qi Wu, Feng Zhang, Jianan Jiang, Ruigao Yang

A modular production-ready image-to-3D generation pipeline coupling SDF Flow-Matching DiT geometry, 3D texture fields, MatWeaver PBR material retrieval, and semantic part decomposition (PartVAE & PartDiT).

Project Page arXiv

BibTeX Citation

@article{fei2026home3d,
  title   = {Home3D 1.0: A High-Fidelity Image-to-3D Asset Generation System for Interior Design},
  author  = {Fei, Yiyun and Li, Guoqiu and Song, Jin and Wu, Chuqiao and Wu, Delong and Wu, Hong and Zeng, Ziru and Chen, Haohui and Kong, Yindong and Li, Jing and Wu, Qi and Zhang, Feng and Jiang, Jianan and Yang, Ruigao},
  journal = {arXiv preprint arXiv:2606.27923},
  year    = {2026},
  url     = {https://homeai.taobao.com/}
}

WIPES: Wavelet-based Visual Primitives ICCV 2025

Wenhao Zhang*, Hao Zhu*, Delong Wu*, Di Kang, Linchao Bao, Xun Cao, Zhan Ma

Representing multi-dimensional visual signals with compact wavelet primitives and a wavelet-based rasterizer for high-frequency 2D image and 5D/6D novel view synthesis.

Project Page arXiv Code

BibTeX Citation

@inproceedings{zhang2025wipes,
  title     = {WIPES: Wavelet-based Visual Primitives},
  author    = {Zhang, Wenhao and Zhu, Hao and Wu, Delong and Kang, Di and Bao, Linchao and Cao, Xun and Ma, Zhan},
  booktitle = {Proceedings of the IEEE/CVF International Conference on Computer Vision (ICCV)},
  year      = {2025}
}

Neural Poisson Solver: A Universal and Continuous Framework for Natural Signal Blending ECCV 2024

Delong Wu*, Hao Zhu*, Qi Zhang, You Li, Zhan Ma, Xun Cao

Solving the Poisson equation directly in continuous implicit neural space for infinite-resolution, boundary-condition-free blending across 2D images and 5D radiance fields.

Project Page arXiv Code

BibTeX Citation

@inproceedings{wu2024neural,
  title     = {Neural Poisson Solver: A Universal and Continuous Framework for Natural Signal Blending},
  author    = {Wu, Delong and Zhu, Hao and Zhang, Qi and Li, You and Ma, Zhan and Cao, Xun},
  booktitle = {European Conference on Computer Vision (ECCV)},
  year      = {2024}
}