Yue Duan

Yue Duan (段岳)

Ph.D. candidate


Affiliation: Reasoning and Learning Group
- School of Computer Science
- National Institute of Healthcare Data Science
- State Key Laboratory for Novel Software Technology
- Nanjing University, Nanjing 210023, China
E-mail: yueduan@smail.nju.edu.cn
Supervisor: Professor Yinghuan Shi

About Me

I am currently a Ph.D. candidate in School of Computer Science at Nanjing University and a member of Reasoning and Learning Group led by Prof. Yang Gao, fortunately supervised by Prof. Yinghuan Shi. My research interest lies in semi-supervised learning, multimodal learning and large multimodal models.

News

  • [Jul 2026] One first-author paper "Unified Multimodal Model Proficient in Continued Training" accepted to ACM Multimedia 2026 — proposes a unified multimodal model capable of continual learning without forgetting.

  • [May 2026] Started as an A-STAR Intern at Alibaba Cloud (ATH-MaaS), working on AI-generated video quality inspection.

  • [Mar 2026] Three co-authored papers accepted to CVPR 2026, AAAI 2026, and ICML 2026 on MLLM hallucination mitigation, vision-language continual learning, and model merging.

Interests
  • Semi-supervised Learning
  • Multimodal Learning
  • Large Multimodal Models
Education

  • Ph.D. candidate in Computer Science and Technology

    Nanjing University, 2023 -

  • M.Sc.'s student as part of the Ph.D. program in Computer Science and Technology

    Nanjing University, 2021 - 2023

  • B.Eng. in Computer Science and Technology

    Harbin Institute of Technology, Weihai, 2017 - 2021

Publications

First-Author Publications

NEWACMMM 2026 Unified Multimodal Model Proficient in Continued Training
ACM International Conference on Multimedia (ACM MM), 2026.
Yue Duan, Yixuan Shi, Zheng Yuan, Zhicheng Zhang, Lei Qi, Chunyu Wang, Yinghuan Shi.
TIP 2025 An Adaptor for Triggering Semi-Supervised Learning to Out-of-Box Serve Deep Image Clustering
IEEE Transactions on Image Processing (TIP), 2025.
Yue Duan, Lei Qi, Yinghuan Shi, Yang Gao.
PDF-arXiv PDF-Published
ICCV 2025 Divide-and-Conquer for Enhancing Unlabeled Learning, Stability, and Plasticity in Semi-supervised Continual Learning
IEEE/CVF International Conference on Computer Vision (ICCV), 2025.
Yue Duan, Taicai Chen, Lei Qi, Yinghuan Shi.
PDF-arXiv PDF-Published Code Video/Poster/Slides 文章解读-知乎
ACMMM 2024 PC$^2$: Pseudo-Classification Based Pseudo-Captioning for Noisy Correspondence Learning in Cross-Modal Retrieval
ACM International Conference on Multimedia (ACM MM), 2024.
Yue Duan, Zhangxuan Gu, Zhenzhe Ying, Lei Qi, Changhua Meng, Yinghuan Shi.
PDF-arXiv PDF-Published Code Dataset Video Poster/Slides 文章解读-知乎 视频解读-bilibili
AAAI 2024 Roll with the Punches: Expansion and Shrinkage of Soft Label Selection for Semi-supervised Fine-Grained Learning
AAAI Conference on Artificial Intelligence (AAAI), 2024.
Yue Duan, Zhen Zhao, Lei Qi, Luping Zhou, Lei Wang, Yinghuan Shi.
PDF-arXiv PDF-Published Code Video Poster/Slides

Co-Authored Publications

NEWCVPR 2026 One Token, Two Fates: A Unified Framework via Vision Token Manipulation Against MLLMs Hallucination
IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 2026.
Zhan Fa, Yue Duan, Jian Zhang, Lei Qi, Yinghuan Shi.
PDF-arXiv PDF-Published Code
NEWAAAI 2026 Decomposing and Composing: Towards Efficient Vision-Language Continual Learning via Rank-1 Expert Pool in a Single LoRA
AAAI Conference on Artificial Intelligence (AAAI), 2026.
Zhan Fa, Yue Duan, Jian Zhang, Lei Qi, Wanqi Yang, Yinghuan Shi.
PDF-arXiv PDF-Published Code Video/Poster/Slides
NEWICML 2026 When Shared Knowledge Hurts: Spectral Over-Accumulation in Model Merging
International Conference on Machine Learning (ICML), 2026.
Yayuan Li, Ze Peng, Jian Zhang, Jintao Guo, Yue Duan, Yinghuan Shi.
PDF-arXiv Code
AAAI 2024 PG-LBO: Enhancing High-Dimensional Bayesian Optimization with Pseudo-Label and Gaussian Process Guidance
AAAI Conference on Artificial Intelligence (AAAI), 2024.
Taicai Chen, Yue Duan, Dong Li, Lei Qi, Yinghuan Shi, Yang Gao.
PDF-arXiv PDF-Published Code Video
CVPR 2022 DC-SSL: Addressing Mismatched Class Distribution in Semi-supervised Learning
IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 2022.
Zhen Zhao, Luping Zhou, Yue Duan, Lei Wang, Lei Qi, Yinghuan Shi.
PDF-Published

Grants

  • NSFC Youth Student Basic Research Program (Ph.D. Candidate): "Theoretical Research and Application of Continual Learning in Multimodal Large Language Models"

    国家自然科学基金委青年学生基础研究项目（博士研究生）：“面向多模态大模型的持续学习理论与方法研究”

    Lead PI300,000￥January 2025 - December 2027

  • 2025 CIE-Tencent Ph.D. Candidate Research Incentive Program (Hunyuan Large Model Special Project): "Research on Visual Generation Large Models", selected as one of only 23 people worldwide

    2025中国电子学会-腾讯博士生科研激励计划（混元大模型专项）：“视觉生成大模型研究”，全球仅23人入选

    Lead PI100,000￥August 2025 - August 2026

Awards & Honors

Student Awards
  • 2025Ministry of Education National Scholarship (Ph.D. Candidates), awarded to 20,000 Ph.D. candidates nationwide

    国家奖学金（博士研究生），全国20,000人

  • 2025Moutai "Pillars of the Nation" Talent Program (Graduate Students), awarded to top 3 graduate students university-wide

    “中国茅台·国之栋梁”本硕博优才计划（研究生），全校仅3人

  • 2024NJU Special Grand Prize of the Pillar Scholarship (Graduate Students), awarded to top 9 graduate students university-wide

    南京大学栋梁奖学金特等奖（研究生），全校仅9人

  • 2023NJU Guorui Scholarship of The 14th Research Institute of China Electronics Technology Group Corporation, 1 recipient in School of Computer Science

    中国电科十四所国睿奖学金，计算机学院仅1人

  • 2022NJU Talent First Prize Scholarship

    南京大学一等英才奖学金

  • 2020HIT Renmin First Prize Scholarship

    哈尔滨工业大学一等人民奖学金

  • 2019HIT Renmin First Prize Scholarship

    哈尔滨工业大学一等人民奖学金

Student Honors
  • 2025Tencent Rhino-bird Research Elite Program, selected as one of only 70+ people worldwide

    腾讯犀牛鸟精英人才计划，全球仅70+人入选
  • 2024NJU Outstanding Graduate Student Role Model, 15 recipients in School of Computer Science

    南京大学优秀研究生标兵，计算机学院共15人

  • 2022NJU Outstanding Graduate Student

    南京大学优秀研究生

  • 2021HIT Outstanding Graduating Students and Thesis Awards

    哈尔滨工业大学优秀毕业生/毕业论文

Contests
  • 2020MCM/ICM Finalist

    美国大学生数学建模竞赛特等奖提名

  • 2019MCM/ICM Honorable Mention

    美国大学生数学建模竞赛二等奖

  • 2019CUMCM Provincial Third Prize

    中国大学生数学建模竞赛省级三等奖

Experience

 
 
 
 
 
Alibaba Group
Alibaba Cloud
ATH-MaaS
A-STAR Intern, Alibaba Cloud
ATH-MaaSHangzhouMay 2026 - Now
AI-generated Video Quality Inspection Model Post-training (AI视频质检模型后训练)
 
 
 
 
 
Tencent
Hunyuan
Rhino-bird Program
Rhino-bird Research Elite Program, Tencent
Hunyuan (混元)ShanghaiJune 2025 - February 2026
Research on Continual Pre-training and Fine-Tuning of Text-to-Image/Video Models (文生图/视频模型的继续预训练和精调研究)
 
 
 
 
 
Ant Group
Ant Security Lab
Research Intern, Ant Group
TiansuanLab (天筭安全实验室)ShanghaiMay 2023 - November 2023
Network-Wide Inspection: Pre-training with Large-Scale Site Data (全网巡检：大规模站点数据预训练)
 
 
 
 
 
Nanjing University
NIHDS of NJU
Research Intern, Nanjing University
National Institute of Healthcare Data Science (健康医疗大数据国家研究院)NanjingMarch 2021 - May 2021
Research on Semi-supervised Medical Image Detection (半监督医疗影像检测算法研究)