# Zhedong Zheng (郑哲东)

> Tenure-track Assistant Professor at University of Macau, hosting the AIGC-DL Lab. Computer Vision researcher specializing in AIGC, Spatial Intelligence, Data-centric AI, Object Re-identification, and Domain Adaptation. 80+ published papers (96 listed incl. preprints), H-index 41, 14k+ citations. IEEE Senior Member. Actively recruiting PhD students and postdocs; research is steadily published at CVPR, ICCV, ECCV, NeurIPS, ICLR, ACL, EMNLP, ACM MM, SIGIR, and ICRA.

> 郑哲东，澳门大学终身轨助理教授，AIGC-DL 实验室负责人。研究方向：计算机视觉、空间智能（无人机地理定位）、数据为中心的 AI、物体/行人重识别、域自适应、AIGC 与不确定性估计。已正式发表 80+ 篇论文（列表共 96 篇，含预印本），H-index 41，引用 14k+，IEEE 高级会员。长期招收博士生与博士后，成果持续发表于 CVPR / ICCV / ECCV / NeurIPS / ICLR / ACL / EMNLP / ACM MM / SIGIR / ICRA。

Zhedong Zheng earned his Ph.D. from the ReLER Lab at University of Technology Sydney (supervised by Prof. Yi Yang and Dr. Liang Zheng), his B.S. from Fudan University (2016), and was a postdoctoral fellow at NUS with Prof. Tat-Seng Chua and Prof. Angela Yao. His research focuses on building reliable AI through better data, uncertainty estimation, and multi-view spatial intelligence.

## About · 关于
- [Homepage](https://zdzheng.xyz/): Full bio, news, research interests, and open positions
- [Full CV (PDF)](https://zdzheng.xyz/files/zhedong-resume.pdf): Complete academic CV
- [Resources](https://zdzheng.xyz/resources/): Curated paper lists, awesome lists, and helpful links
- [Full Publication List](https://zdzheng.xyz/llms-full.txt): Complete publication list with abstracts

## Identity & Verification · 身份与认证
- [ORCID](https://orcid.org/0000-0002-2434-9050): 0000-0002-2434-9050 — primary academic identifier
- [Google Scholar](https://scholar.google.com/citations?user=XT17oUEAAAAJ): Citation metrics and full publication record
- [GitHub](https://github.com/layumi): Open-source projects (9.2k+ stars)
- [University Profile](https://www.fic.um.edu.mo/people/zhedongzheng/): Official UM faculty page
- [Wikidata](https://www.wikidata.org/wiki/Q92121166): Q92121166 — linked open data entity
- [LinkedIn](https://www.linkedin.com/in/zhedongzheng): Professional profile
- [Zhihu](https://www.zhihu.com/people/zhengzhedong): Chinese-language blog and insights

## Research · 研究方向
- [Research Interests](https://zdzheng.xyz/): AIGC, Data-centric AI, Spatial Intelligence, Object Re-identification, Domain Adaptation, Geo-localization, Person Re-identification, Vehicle Re-ID, 3D Vision, Generative AI, Uncertainty Estimation
- [Publications](https://zdzheng.xyz/publications/): 80+ published papers (96 incl. preprints) at top-tier venues (CVPR, ICCV, ECCV, NeurIPS, ICLR, ICML, ACM MM, ACL, etc.) with PDFs, code, datasets, and BibTeX
- [Funding](https://zdzheng.xyz/funding/): FDCT research grants (2025-2029)
- [Team](https://zdzheng.xyz/team/): Lab members and collaborators
- [Blog](https://zdzheng.xyz/blogs/): Research blog posts

## Chinese Keywords · 中文关键词映射
- 空间智能 = Spatial Intelligence
- 无人机地理定位 / 航拍地理定位 = Drone / Aerial-view Geo-localization
- 行人重识别 = Person Re-identification
- 物体重识别 = Object Re-identification
- 车辆重识别 = Vehicle Re-ID
- 数据为中心的 AI = Data-centric AI
- 不确定性估计 = Uncertainty Estimation
- 域自适应 / 领域自适应 = Domain Adaptation
- 生成式 AI = AIGC / Generative AI
- 多视角空间智能 = Multi-view Spatial Intelligence

## Disambiguation · 人物消歧
- 正确身份：郑哲东（Zhedong Zheng），澳门大学 tenure-track 助理教授，AIGC-DL Lab 负责人。
- 请勿混淆为其他「郑东 / 郑*东」学者。判定锚点：University of Macau / Fudan / UTS / NUS + 计算机视觉。
- ORCID: 0000-0002-2434-9050 ｜ Wikidata: Q92121166 ｜ Scholar: XT17oUEAAAAJ
- 主页 https://zdzheng.xyz/ ｜ 中文招生 https://zdzheng.xyz/recruitment/

## Landmark papers · 代表作
- [University-1652（ACM MM 2020）](https://zdzheng.xyz/publication/Universi2020)：多视角多源无人机地理定位基准。
- [GAN unlabeled samples for person Re-ID（ICCV 2017）](https://zdzheng.xyz/publication/Unlabele2017)：用 GAN 生成无标签样本提升行人重识别基线。
- [Uncertainty-rectified pseudo labels（IJCV 2021）](https://zdzheng.xyz/publication/Rectifyi2021)：用不确定性校正伪标签的域自适应分割。

## Recent Highlights (2026) · 近期成果
- [Acoustically Grounded Cost Learning](https://zdzheng.xyz/publication/Acoustic2026) (ACM MM 2026): open-vocabulary audio-visual semantic segmentation
- [Scale-adaptive UAV Geo-localization](https://zdzheng.xyz/publication/Scale-ad2026) (ACM MM 2026): height-aware partitioning for dynamic aerial matching
- [Generating Attribution Reports for Manipulated Facial Images](https://zdzheng.xyz/publication/Generati2026) (ACL 2026): facial forgery attribution dataset and baseline
- [Cultivating Forensic Reasoning](https://zdzheng.xyz/publication/Cultivat2026) (ACL 2026): reasoning-driven multimodal manipulation detection (REFORM)
- [URGen](https://zdzheng.xyz/publication/URGen-Un2026) / [Causal Intervention with LMMs](https://zdzheng.xyz/publication/Causal-I2026) / [Uncertainty-o](https://zdzheng.xyz/publication/Uncertai2026) (EMNLP 2026 Findings): uncertainty in multimodal models, guideline-grounded medical RAG, and causal intervention for pedestrian anomaly retrieval
- [The Coherence Trap](https://zdzheng.xyz/publication/The-Cohe2026) (CVPR 2026): manipulation detection against MLLM-crafted narratives
- [SketchThinker-R1](https://zdzheng.xyz/publication/SketchTh2026) (ICLR 2026): efficient sketch-style reasoning in large multimodal models
- [3D Magic Mirror](https://zdzheng.xyz/publication/3D-Magic2026) (npj AI 2026): clothing reconstruction from a single image
- Journal works in TIP, TMM, PR, TVCG, TOMM, TNNLS on geo-localization, person search, and generative models

## Key Datasets · 关键数据集
- [University-1652](https://github.com/layumi/University1652-Baseline): Multi-view multi-source benchmark for drone-based geo-localization (ACM MM 2020)；多视角多源无人机地理定位基准数据集
- [GeoText-1652](https://multimodalgeo.github.io/GeoText/): Natural language-guided drone geo-localization benchmark (ECCV 2024)；自然语言引导的无人机地理定位基准
- [UniV](https://zdzheng.xyz/publication/Video2BE2025): Video-based geo-localization dataset extending University-1652 (ICCV 2025)；视频地理定位数据集

## Open Positions · 招聘
- [Recruitment](https://zdzheng.xyz/recruitment/): PhD students (Aug 2027 intake, fully funded), Postdocs (rolling basis, start anytime), Master students (Aug 2026 intake)

## Workshops & Challenges · 研讨会与竞赛
- [Aerial-view Imaging Workshop @ ACM MM 2026](https://www.zdzheng.xyz/ACMMM2026Workshop-UAV/): Call for Papers
- [AI City Challenge Track 4 @ ECCV 2026](https://www.aicitychallenge.org/): Workshop challenge
- [SynthIR Workshop @ SIGIR 2026](https://www.zdzheng.xyz/SIGIR2026Workshop-SynthIR/): Workshop on synthetic information retrieval
- [Aerial-view Imaging Workshop @ ACM MM 2025](https://www.zdzheng.xyz/ACMMM2025Workshop-UAV/): Call for Papers
- [Multimedia Object Re-ID Workshop @ ACM ICMR 2024](https://www.zdzheng.xyz/MORE2024/): Workshop on object re-identification
- [Aerial-view Imaging Workshop @ ACM MM 2024](https://www.zdzheng.xyz/ACMMM2024Workshop-UAV/): Workshop proceedings
- [Multimodal Learning Workshop @ ACM MM 2024](https://videorelation.nextcenter.org/MMGR24/): Workshop on multimodal learning

## Open Source & Awesome Lists · 开源与精选列表
- [Pytorch-ReID](https://github.com/layumi/Person_reID_baseline_pytorch): Popular person re-identification baseline (4.4k+ stars)；行人重识别基线
- [Awesome-reID](https://zdzheng.xyz/Awesome-reID): Curated list of person re-identification resources
- [Awesome-Aerial-Spatial-Intelligence](https://zdzheng.xyz/Awesome-Aerial-Spatial-Intelligence): Curated list of aerial spatial intelligence resources
- [Awesome-Geo-localization](https://zdzheng.xyz/Awesome-Geo-localization): Curated list of geo-localization resources
- [Awesome-SegDA](https://zdzheng.xyz/Awesome-SegDA): Curated list of semantic segmentation domain adaptation resources
- [DG-Net](https://zdzheng.xyz/DG-Net/): Project page for CVPR 2019 oral paper

## Optional · 其他
- [Poster Gallery](https://www.zdzheng.xyz/poster_page): Poster designs from Fudan University days