Zitai Wang

Zitai Wang

Post-doc at Institute of Computing Technology (ICT), CAS

About Me

Hi! I’m Zitai Wang (王子泰, E-mail: wangzitai@ict.ac.cn). Now I am a Post-doc Fellow at Institute of Computing Technology, Chinese Academy of Sciences (ICT, CAS). I received the Ph.D. degree in Institute of Information Engineering, Chinese Academy of Sciences (IIE, CAS), supervised by Prof. Qingming Huang (黄庆明) (IEEE Fellow). I am also lucky to have opportunities to collaborate with Qianqian Xu (许倩倩) (Professor at ICT, CAS), Xiaochun Cao (操晓春) (Professor at Shenzhen Campus of Sun Yat-sen University), Zhiyong Yang (杨智勇) (Associate Professor at UCAS), Ke Ma (马坷) (Associate Professor at UCAS), Yangbangyan Jiang (姜阳邦彦) (Postdoc at UCAS), Shilong Bao (包世龙) (Postdoc at UCAS), and Peisong Wen (温佩松) (Postdoc at UCAS).

My research interest includes machine learning and data mining. I have authored or co-authored 20+ academic papers in top-tier international conferences and journals, including TPAMI, IJCV, ICML, NeurIPS, AAAI, and ACM Multimedia.

王子泰 (邮箱：wangzitai@ict.ac.cn)，现为中国科学院计算技术研究所博士后/特别研究助理 (合作导师：许倩倩研究员)，博士毕业于中国科学院信息工程研究所（导师：黄庆明教授）。主要研究方向为机器学习与数据挖掘，已在 CCF-A 类期刊/会议发表论文 20 余篇，其中TPAMI/IJCV/ICML/NeurIPS 10 余篇，并有 7 篇入选 Oral/Spotlight 论文，获 CVPR 国际竞赛冠军；入选中国图象图形学学会博士学位论文激励计划（CSIG优博）、ACM北京分会优秀博士学位论文奖、中国科学院院长优秀奖；获博士后创新人才支持计划（博新计划）、国家自然科学基金青年项目C类、北京市自然科学基金-海淀原始创新联合基金培育项目、中国科学院特别研究助理资助项目、中国博士后科学基金面上项目资助；担任 TPAMI、ICML、NeurIPS 等 CCF-A 类期刊/会议审稿人，被评为 NeurIPS Top 审稿人。

📝 Publications

2026

Quantifying the Potential to Escape Filter Bubbles: A Behavior-Aware Measure via Contrastive Simulation

Difu feng, Qianqian Xu, Zitai Wang, Cong Hua, Zhiyong Yang and Qingming Huang. Quantifying the Potential to Escape Filter Bubbles: A Behavior-Aware Measure via Contrastive Simulation. AAAI Conference on Artificial Intelligence (AAAI), 2026. (Accepted) | [Code]

2025

A Unified Perspective for Loss-Oriented Imbalanced Learning via Localization

Zitai Wang, Qianqian Xu, Zhiyong Yang, Zhikang Xu, Linchao Zhang, Xiaochun Cao and Qingming Huang. A Unified Perspective for Loss-Oriented Imbalanced Learning via Localization. IEEE Transactions on Pattern Analysis and Machine Intelligence (TPAMI, IF: 18.6), 48(1): 639-656, Jan. 2026. | [Code]

DirMixE: Harnessing Test Agnostic Long-tail Recognition with Hierarchical Label Variations

Zhiyong Yang, Qianqian Xu, Sicong Li, Zitai Wang, Xiaochun Cao and Qingming Huang. DirMixE: Harnessing Test Agnostic Long-tail Recognition with Hierarchical Label Variations. IEEE Transactions on Pattern Analysis and Machine Intelligence (TPAMI, IF: 18.6), 2025. (Accepted, DOI: 10.1109/TPAMI.2025.3647124)

ABKD: Pursuing a Proper Allocation of the Probability Mass in Knowledge Distillation via $\alpha$-$\beta$-Divergence

Guanghui Wang, Zhiyong Yang, Zitai Wang, Shi Wang, Qianqian Xu and Qingming Huang. ABKD: Pursuing a Proper Allocation of the Probability Mass in Knowledge Distillation via $\alpha$-$\beta$-Divergence. International Conference on Machine Learning (ICML), 65167-65212, 2025. (Oral, 1.0%) | [Code]

OpenworldAUC: Towards Unified Evaluation and Optimization for Open-world Prompt Tuning

Cong Hua, Qianqian Xu, Zhiyong Yang, Zitai Wang, Shilong Bao and Qingming Huang. OpenworldAUC: Towards Unified Evaluation and Optimization for Open-world Prompt Tuning. International Conference on Machine Learning (ICML), 24975-25020, 2025. [Code]

Focal-SAM: Focal Sharpness-Aware Minimization for Long-Tailed Classification

Sicong Li, Qianqian Xu, Zhiyong Yang, Zitai Wang, Linchao Zhang, Xiaochun Cao and Qingming Huang. Focal-SAM: Focal Sharpness-Aware Minimization for Long-Tailed Classification. International Conference on Machine Learning (ICML), 36624-36651, 2025. | [Code]

EDGE: Unknown-aware Multi-label Learning by Energy Distribution Gap Expansion

Yuchen Sun, Qianqian Xu, Zitai Wang, Zhiyong Yang and Junwei He. EDGE: Unknown-aware Multi-label Learning by Energy Distribution Gap Expansion. AAAI Conference on Artificial Intelligence (AAAI), 12613-12621, 2025. | [Code]

2024

Top-K Pairwise Ranking: Bridging the Gap Among Ranking-Based Measures for Multi-Label Classification

Zitai Wang, Qianqian Xu, Zhiyong Yang, Peisong Wen, Yuan He, Xiaochun Cao and Qingming Huang. Top-K Pairwise Ranking: Bridging the Gap Among Ranking-Based Measures for Multi-Label Classification. International Journal of Computer Vision (IJCV, IF: 9.3), 133(1): 211-253, Jan. 2025. | [Code]

Harnessing Hierarchical Label Distribution Variations in Test Agnostic Long-tail Recognition

Zhiyong Yang, Qianqian Xu, Zitai Wang, Sicong Li, Boyu Han, Shilong Bao, Xiaochun Cao and Qingming Huang. Harnessing Hierarchical Label Distribution Variations in Test Agnostic Long-tail Recognition. International Conference on Machine Learning (ICML), 56624-56664, 2024. | [Code] | [Video]

Not All Diffusion Model Activations Have Been Evaluated as Discriminative Features

Benyuan Meng, Qianqian Xu, Zitai Wang, Xiaochun Cao and Qingming Huang. Not All Diffusion Model Activations Have Been Evaluated as Discriminative Features. Annual Conference on Neural Information Processing Systems (NeurIPS), 55141-55177, 2024. (Spotlight, 2.1%) | [Code]

Suppress Content Shift: Better Diffusion Features via Off-the-Shelf Generation Techniques

Benyuan Meng, Qianqian Xu, Zitai Wang, Zhiyong Yang, Xiaochun Cao and Qingming Huang. Suppress Content Shift: Better Diffusion Features via Off-the-Shelf Generation Techniques. Annual Conference on Neural Information Processing Systems (NeurIPS), 18910-18939, 2024. [Code]

ADA-GAD: Anomaly-Denoised Autoencoders for Graph Anomaly Detection

Junwei He, Qianqian Xu, Yangbangyan Jiang, Zitai Wang and Qingming Huang. ADA-GAD: Anomaly-Denoised Autoencoders for Graph Anomaly Detection. AAAI Conference on Artificial Intelligence (AAAI), 8481-8489, 2024. | [Code]

HGOE: Hybrid External and Internal Graph Outlier Exposure for Graph Out-of-Distribution Detection

Junwei He, Qianqian Xu, Yangbangyan Jiang, Zitai Wang, Yuchen Sun and Qingming Huang. HGOE: Hybrid External and Internal Graph Outlier Exposure for Graph Out-of-Distribution Detection. ACM Conference on Multimedia (ACM MM), 1544-1553, 2024. | [Code]

Before

Optimizing Partial Area Under the Top-k Curve: Theory and Practice

Zitai Wang, Qianqian Xu, Zhiyong Yang, Yuan He, Xiaochun Cao and Qingming Huang. Optimizing Partial Area Under the Top-k Curve: Theory and Practice. IEEE Transactions on Pattern Analysis and Machine Intelligence (TPAMI, IF: 18.6), 45(4): 5053-5069, Apr. 2023. | [Code]

A Unified Generalization Analysis of Re-Weighting and Logit-Adjustment for Imbalanced Learning

Zitai Wang, Qianqian Xu, Zhiyong Yang, Yuan He, Xiaochun Cao and Qingming Huang. A Unified Generalization Analysis of Re-Weighting and Logit-Adjustment for Imbalanced Learning. Annual Conference on Neural Information Processing Systems (NeurIPS), 48417-48430, 2023. (Spotlight, 3.1%) | [Code] | [Video] | [Slides]

OpenAUC: Towards AUC-Oriented Open-Set Recognition

Zitai Wang, Qianqian Xu, Zhiyong Yang, Yuan He, Xiaochun Cao and Qingming Huang. OpenAUC: Towards AUC-Oriented Open-Set Recognition. Annual Conference on Neural Information Processing Systems (NeurIPS), 25033-25045, 2022. (Spotlight, 5.4%) | [Code] | [Video] | [Poster] | [Slides]

Confederated Learning: Going Beyond Centralization

Zitai Wang, Qianqian Xu, Ke Ma, Xiaochun Cao and Qingming Huang. Confederated Learning: Going Beyond Centralization. ACM Conference on Multimedia (ACM MM), 2939-2947, 2022. (Oral, 5.6%) | [Code] | [Slides]

Implicit feedbacks are not always favorable: Iterative Relabeled One-Class Collaborative Filtering against Noisy Interactions

Zitai Wang, Qianqian Xu, Zhiyong Yang, Xiaochun Cao and Qingming Huang. Implicit feedbacks are not always favorable: Iterative Relabeled One-Class Collaborative Filtering against Noisy Interactions. ACM Conference on Multimedia (ACM MM), 3070-3078, 2021. | [Code] | [Poster] | [Slides]

Adversarial Preference Learning with Pairwise Comparisons

Zitai Wang, Qianqian Xu, Ke Ma, Yangbangyan Jiang, Xiaochun Cao and Qingming Huang. Adversarial Preference Learning with Pairwise Comparisons. ACM Conference on Multimedia (ACM MM), 656-664, 2019. (Oral, 9.4%) | [Code] | [Poster] | [Slides]

When Measures are Unreliable: Imperceptible Adversarial Perturbations toward Top-k Multi-Label Learning

Yuchen Sun, Qianqian Xu, Zitai Wang, and Qingming Huang. When Measures are Unreliable: Imperceptible Adversarial Perturbations toward Top-k Multi-Label Learning. ACM Conference on Multimedia (ACM MM), 1515-1526, 2023. (Oral, 5.4%) | [Code]

🧑‍🎓 Student Collaborators

Ph.D. Students

  • Cong Hua (supervisor: Qingming Huang) (华聪，导师：黄庆明): ICML 2025, ICML 2024（中国科协青年科技人才培育工程博士生专项计划）
  • Sicong Li (supervisor: Qingming Huang) (李斯骢，导师：黄庆明): ICML 2025

Master Students

  • Yuchen Sun (supervisor: Qianqian Xu) (孙宇辰，导师：许倩倩): AAAI 2025, ACM MM 2023 (Oral), 计算机学报（中国图象图形学学会优硕、北京图象图形学学会优硕）
  • Junwei He (supervisor: Qingming Huang) (何俊伟，导师：黄庆明): ACM MM 2024, AAAI 2024（国家奖学金）
  • Benyuan Meng (supervisor: Qingming Huang) (孟本源，导师：黄庆明): NeurIPS 2024 (Spotlight), NeurIPS 2024
  • Difu Feng (supervisor: Qingming Huang) (冯迪夫，导师：黄庆明): AAAI 2026
  • Guanghui Wang (supervisor: Zhiyong Yang) (王广辉，导师：杨智勇): ICML 2025 (Oral)

🎖 Honors & Awards

  • 2025 CSIG Outstanding Doctoral Dissertation Award (中国图象图形学学会优博)
  • 2025 ACM Beijing Doctorial Dissertation Award (ACM北京分会优秀博士论文奖)
  • 2025 1st Place Award at the 3rd CVPR Workshop on Compositional 3D Vision (Coarse-Grained GCR Track)
  • 2025 ICLR 2025 Notabe Reviewer (480/all) (ICLR 2025 Notabe 审稿人)
  • 2024 NeurIPS 2024 Top Reviewer (1304/15160, 8.6%) (NeurIPS 2024 Top 审稿人)
  • 2024 CAS President Award (Excellent Prize) (中国科学院院长优秀奖)
  • 2024 Excellent Graduate of Beijing & UCAS. (北京市优秀毕业生，中国科学院大学优秀毕业生)
  • 2023 National Scholarship, Ministry of Education of the PRC. (国家奖学金)
  • 2021 Director Scholarship (Excellent Prize), IIE, CAS. (中国科学院信工所所长优秀奖)
  • 2019 Excellent Graduate of Beijing & BJTU. (北京市优秀毕业生，北京交通大学优秀毕业生)

💻 Projects & Fundings

  • 2025 Young Scientists Fund (C Class) of the National Natural Science Foundation of China (国家自然科学基金青年科学基金C类)
  • 2025 General Program of the Chinese Postdoctoral Science Foundation (中国博士后科学基金面上资助)
  • 2025 Incubation Project of the Beijing Natural Science Foundation – Haidian Original Innovation Joint Fund (北京市自然科学基金-海淀原始创新联合基金培育项目)
  • 2025 Special Research Assistant Funding Program of the Chinese Academy of Sciences (中国科学院特别研究助理资助项目)
  • 2024 Postdoctoral Innovative Talents Support Program (博士后创新人才支持计划)

💬 Invited Talks & Presentations

  • 2025.11:   Beijing Youth Science and Technology Salon (北京青年科技沙龙). [Page].
  • 2024.01:   TechBeat Talk of NeurIPS 2023. [Video].
  • 2023.12:   AI TIME Talk of NeurIPS 2023. [Page].
  • 2023.12:   Young Scientists Conference of CSIG (CSIG青年科学家会议). [Poster].
  • 2023.10:   PhD. student Forum of PRCV 2023 (PRCV博士生论坛). [Page].
  • 2023.02:   AI TIME Youth PhD Talk of NeurIPS 2022. [Video].

📖 Academic Services

Journal

  • IEEE Transactions on Pattern Analysis and Machine Intelligence (IEEE TPAMI): Reviewer
  • IEEE Transactions on Image Processing (IEEE TIP): Reviewer
  • IEEE Transactions on Knowledge and Data Engineering (IEEE TKDE): Reviewer
  • IEEE Transactions on Circuits and Systems for Video Technology (IEEE TCSVT): Reviewer
  • IEEE Transactions on Neural Networks and Learning Systems (IEEE TNNLS): Reviewer
  • IEEE Transactions on Multimedia (IEEE TMM): Reviewer
  • Transactions on Machine Learning Research (TMLR): Reviewer
  • Pattern Recognition (PR): Reviewer
  • Machine Learning (ML): Reviewer
  • CAAI Transactions on Intelligence Technology (CAAI TIT): Reviewer

Conference

  • ICML: PC Member (2025, 2026)
  • NeurIPS: PC Member (2024, 2025)
  • ICLR: PC Member (2025, 2026)
  • CVPR: PC Member (2024, 2025, 2026)
  • ICCV: PC Member (2025)
  • AISTATS: PC Member (2025, 2026)
  • AAAI: PC Member (2023, 2024, 2025, 2026)
  • IJCAI: PC Member (2025)
  • ACM MM: PC Member (2023, 2024, 2025)
  • ICDE: PC Member (2024)
  • ACML: PC Member (2024, 2025)