Bo Han


Home


Bo Han

Assistant Professor in Machine Learning @ Department of Computer Science
Hong Kong Baptist University Faculty of Science

BAIHO Visiting Scientist @ Imperfect Information Learning Team
RIKEN Center for Advanced Intelligence Project

[Google Scholar] [Github] [Group Website]
E-mail: bhanml@comp.hkbu.edu.hk & bo.han@a.riken.jp
TMLR Group is always looking for highly self-motivated PhD/RA/Visiting students and Postdoc researchers. Please read this document for recruiting information, and check this document for department information. Meanwhile, TMLR Group is happy to host remote research trainees. Due to the large number of emails we receive, we cannot respond to every email individually. Thanks!


News

See more news here.


Research

    My research interests lie in machine learning and deep learning. My long-term goal is to develop trustworthy intelligent systems, which can learn and reason from a massive volume of complex (e.g., weakly supervised, self-supervised, robust, causal, fair, and privacy-preserving) data (e.g, label, example, preference, domain, similarity, graph, and demonstration) automatically and federatively. Recently, I develop core machine learning methodology. Besides, I am actively applying our fundamental research into the interdisciplinary domain. I am heading Trustworthy Machine Learning and Reasoning Group.
    My current research work center around four major themes (with representative works/projects; more information here):


Selected Projects

  • RGC ECS (PI): Trustworthy Deep Learning from Open-set Corrupted Data [Link] [Website]

  • NSFC YSF (PI): The Research on the Automated Trustworthy Machine Learning [Link] [Website]

  • GDST BRF (PI): Trustworthy Deep Reasoning with Human-level Constraints

  • RIKEN CRF (PI): New Directions in Trustworthy Machine Learning


Research Highlights

(* indicates advisees/co-advisees; updating through the time; see the full list here)
  • Combating Exacerbated Heterogeneity for Robust Models in Federated Learning.
    J. Zhu*, J. Yao, T. Liu, Q. Yao, J. Xu, and B. Han.
    In Proceedings of 11th International Conference on Learning Representations (ICLR'23), [PDF] [Code] [Poster].

  • Latent Class-Conditional Noise Model.
    J. Yao, B. Han, Z. Zhou, Y. Zhang, and I.W. Tsang.
    IEEE Transactions on Pattern Analysis and Machine Intelligence (PAMI), 2023, [PDF] [Code].

  • Watermarking for Out-of-distribution Detection.
    Q. Wang*, F. Liu, Y. Zhang, J. Zhang, C. Gong, T. Liu, and B. Han.
    In Advances in Neural Information Processing Systems 35 (NeurIPS'22), [PDF] [Code] [Poster] [Spotlight].

  • Is Out-of-distribution Detection Learnable?
    Z. Fang, S. Li, J. Lu, J. Dong, B. Han, and F. Liu.
    In Advances in Neural Information Processing Systems 35 (NeurIPS'22), [PDF] [Poster] [Oral, Outstanding Paper Award].

  • CausalAdv: Adversarial Robustness through the Lens of Causality.
    Y. Zhang*, M. Gong, T. Liu, G. Niu, X. Tian, B. Han, B. Schölkopf, and K. Zhang.
    In Proceedings of 10th International Conference on Learning Representations (ICLR'22), [PDF] [Code] [Poster].

  • Confidence Scores Make Instance-dependent Label-noise Learning Possible.
    A. Berthon*, B. Han, G. Niu, T. Liu, and M. Sugiyama.
    In Proceedings of 38th International Conference on Machine Learning (ICML'21), [PDF] [Code] [Poster] [Long Oral].

  • How does Disagreement Help Generalization against Label Corruption?
    X. Yu*, B. Han, J. Yao, G. Niu, I.W. Tsang, and M. Sugiyama.
    In Proceedings of 36th International Conference on Machine Learning (ICML'19), [PDF] [Code] [Poster] [Long Oral].

  • Towards Robust ResNet: A Small Step but A Giant Leap.
    J. Zhang*, B. Han, L. Wynter, B. Low, and M. Kankanhalli.
    In Proceedings of 28th International Joint Conference on Artificial Intelligence (IJCAI'19), [PDF] [Code] [Poster].

  • Co-teaching: Robust Training of Deep Neural Networks with Extremely Noisy Labels.
    B. Han, Q. Yao, X. Yu, G. Niu, M. Xu, W. Hu, I.W. Tsang, and M. Sugiyama.
    In Advances in Neural Information Processing Systems 31 (NeurIPS'18), [PDF] [Code] [Poster] [Most Influential Paper Award].

  • Masking: A New Perspective of Noisy Supervision.
    B. Han, J. Yao, G. Niu, M. Zhou, I.W. Tsang, Y. Zhang, and M. Sugiyama.
    In Advances in Neural Information Processing Systems 31 (NeurIPS'18), [PDF] [Code] [Poster].


Selected Publications

(* indicates advisees/co-advisees; see the full list here)
  • Machine Learning with Noisy Labels: From Theory to Heuristics.
    M. Sugiyama, N. Lu, B. Han, T. Liu, and G. Niu.
    Adaptive Computation and Machine Learning series, The MIT Press, 2024, [PDF].
    (the monograph is accepted; coming in 2024)

  • A Survey of Label-noise Representation Learning: Past, Present and Future.
    B. Han, Q. Yao, T. Liu, G. Niu, I.W. Tsang, J.T. Kwok, and M. Sugiyama.
    arXiv preprint arXiv:2011.04406, 2020, [PDF].
    (the draft is kept updating; any comments and suggestions are welcome)

  • Combating Exacerbated Heterogeneity for Robust Models in Federated Learning.
    J. Zhu*, J. Yao, T. Liu, Q. Yao, J. Xu, and B. Han.
    In Proceedings of 11th International Conference on Learning Representations (ICLR'23), [PDF] [Code] [Poster].

  • A Holistic View of Noise Transition Matrix in Deep Learning and Beyond.
    Y. Lin*, R. Pi*, W. Zhang, X. Xia, J. Gao, X. Zhou, T. Liu, and B. Han.
    In Proceedings of 11th International Conference on Learning Representations (ICLR'23), [PDF] [Code] [Poster] [Spotlight].

  • Latent Class-Conditional Noise Model.
    J. Yao, B. Han, Z. Zhou, Y. Zhang, and I.W. Tsang.
    IEEE Transactions on Pattern Analysis and Machine Intelligence (PAMI), 2023, [PDF] [Code].

  • KRADA: Known-region-aware Domain Alignment for Open-set Domain Adaptation in Semantic Segmentation.
    C. Zhou*, F. Liu, C. Gong, R. Zeng, T. Liu, W.K. Cheung, and B. Han.
    Transactions on Machine Learning Research (TMLR), 2023, [PDF] [Code].

  • Watermarking for Out-of-distribution Detection.
    Q. Wang*, F. Liu, Y. Zhang, J. Zhang, C. Gong, T. Liu, and B. Han.
    In Advances in Neural Information Processing Systems 35 (NeurIPS'22), [PDF] [Code] [Poster] [Spotlight].

  • Is Out-of-distribution Detection Learnable?
    Z. Fang, S. Li, J. Lu, J. Dong, B. Han, and F. Liu.
    In Advances in Neural Information Processing Systems 35 (NeurIPS'22), [PDF] [Poster] [Oral, Outstanding Paper Award].

  • Virtual Homogeneity Learning: Defending against Data Heterogeneity in Federated Learning.
    Z. Tang*, Y. Zhang, S. Shi, X. He, B. Han, and X. Chu.
    In Proceedings of 39th International Conference on Machine Learning (ICML'22), [PDF] [Code] [Poster].

  • Contrastive Learning with Boosted Memorization.
    Z. Zhou*, J. Yao, Y. Wang, B. Han, and Y. Zhang.
    In Proceedings of 39th International Conference on Machine Learning (ICML'22), [PDF] [Code] [Poster].

  • Understanding and Improving Graph Injection Attack by Promoting Unnoticeability.
    Y. Chen*, H. Yang, Y. Zhang, K. Ma, T. Liu, B. Han, and J. Cheng.
    In Proceedings of 10th International Conference on Learning Representations (ICLR'22), [PDF] [Code] [Poster].

  • CausalAdv: Adversarial Robustness through the Lens of Causality.
    Y. Zhang*, M. Gong, T. Liu, G. Niu, X. Tian, B. Han, B. Schölkopf, and K. Zhang.
    In Proceedings of 10th International Conference on Learning Representations (ICLR'22), [PDF] [Code] [Poster].

  • Bilateral Dependency Optimization: Defending Against Model-inversion Attacks.
    X. Peng*, F. Liu, J. Zhang, J. Ye, L. Lan, T. Liu, and B. Han.
    In Proceedings of 28th ACM Conference on Knowledge Discovery and Data Mining (KDD'22), [PDF] [Code] [Poster].

  • Device-Cloud Collaborative Recommendation via Meta Controller.
    J. Yao, F. Wang, X. Ding, S. Chen, B. Han, J. Zhou, and H. Yang.
    In Proceedings of 28th ACM Conference on Knowledge Discovery and Data Mining (KDD'22), [PDF] [Poster].

  • Fair Classification with Instance-dependent Label Noise.
    S. Wu, M. Gong, B. Han, Y. Liu, and T. Liu.
    In Proceedings of 1st Conference on Causal Learning and Reasoning (CLeaR'22), [PDF] [Code].

  • Low-rank Tensor Learning with Nonconvex Overlapped Nuclear Norm Regularization.
    Q. Yao, Y. Wang, B. Han, and J.T. Kwok.
    Journal of Machine Learning Research (JMLR), 2022, [PDF] [Code].

  • Learning with Mixed Open-set and Closed-set Noisy Labels.
    X. Xia, B. Han, N. Wang, J. Deng, J. Li, Y. Mao, and T. Liu.
    IEEE Transactions on Pattern Analysis and Machine Intelligence (PAMI), 2022, [PDF] [Code].

  • NoiLin: Improving Adversarial Training and Correcting Stereotype of Noisy Labels.
    J. Zhang, X. Xu, B. Han, T. Liu, L. Cui, G. Niu, and M. Sugiyama.
    Transactions on Machine Learning Research (TMLR), 2022, [PDF] [Code].

  • TOHAN: A One-step Approach towards Few-shot Hypothesis Adaptation.
    H. Chi*, F. Liu, W. Yang, L. Lan, T. Liu, B. Han, W.K. Cheung, and J.T. Kwok.
    In Advances in Neural Information Processing Systems 34 (NeurIPS'21), [PDF] [Code] [Poster] [Spotlight].

  • Instance-dependent Label-noise Learning under a Structural Causal Model.
    Y. Yao, T. Liu, M. Gong, B. Han, G. Niu, and K. Zhang.
    In Advances in Neural Information Processing Systems 34 (NeurIPS'21), [PDF] [Code] [Poster].

  • Confidence Scores Make Instance-dependent Label-noise Learning Possible.
    A. Berthon*, B. Han, G. Niu, T. Liu, and M. Sugiyama.
    In Proceedings of 38th International Conference on Machine Learning (ICML'21), [PDF] [Code] [Poster] [Long Oral].

  • Maximum Mean Discrepancy is Aware of Adversarial Attacks.
    R. Gao*, F. Liu, J. Zhang, B. Han, T. Liu, G. Niu, and M. Sugiyama.
    In Proceedings of 38th International Conference on Machine Learning (ICML'21), [PDF] [Code] [Poster].

  • Geometry-aware Instance-reweighted Adversarial Training.
    J. Zhang, J. Zhu*, G. Niu, B. Han, M. Sugiyama, and M. Kankanhalli.
    In Proceedings of 9th International Conference on Learning Representations (ICLR'21), [PDF] [Code] [Poster] [Oral].

  • Robust Early-learning: Hindering the Memorization of Noisy Labels.
    X. Xia, T. Liu, B. Han, C. Gong, N. Wang, Z. Ge, and Y. Chang.
    In Proceedings of 9th International Conference on Learning Representations (ICLR'21), [PDF] [Code] [Poster].

  • Learning with Group Noise.
    Q. Wang*, J. Yao, C. Gong, T. Liu, M. Gong, H. Yang, and B. Han.
    In Proceedings of 35th AAAI Conference on Artificial Intelligence (AAAI'21), [PDF] [Code].

  • Device-Cloud Collaborative Learning for Recommendation.
    J. Yao, F. Wang, K. Jia, B. Han, J. Zhou, and H. Yang.
    In Proceedings of 27th ACM Conference on Knowledge Discovery and Data Mining (KDD'21), [PDF] [Poster].

  • Provably Consistent Partial-Label Learning.
    L. Feng, J. Lv, B. Han, M. Xu, G. Niu, X. Geng, B. An, and M. Sugiyama.
    In Advances in Neural Information Processing Systems 33 (NeurIPS'20), [PDF] [Code] [Poster].

  • Reducing Estimation Error for Transition Matrix in Label-noise Learning.
    Y. Yao, T. Liu, B. Han, M. Gong, J. Deng, G. Niu, and M. Sugiyama.
    In Advances in Neural Information Processing Systems 33 (NeurIPS'20), [PDF] [Code] [Poster].

  • SIGUA: Forgetting May Make Learning with Noisy Labels More Robust.
    B. Han, G. Niu, X. Yu, Q. Yao, M. Xu, I.W. Tsang, and M. Sugiyama.
    In Proceedings of 37th International Conference on Machine Learning (ICML'20), [PDF] [Code] [Poster].

  • Variational Imitation Learning from Diverse-quality Demonstrations.
    V. Tangkaratt, B. Han, M. Khan, and M. Sugiyama.
    In Proceedings of 37th International Conference on Machine Learning (ICML'20), [PDF] [Code] [Poster].

  • Attacks Which Do Not Kill Training Make Adversarial Learning Stronger.
    J. Zhang*, X. Xu, B. Han, G. Niu, L. Cui, M. Sugiyama, and M. Kankanhalli.
    In Proceedings of 37th International Conference on Machine Learning (ICML'20), [PDF] [Code] [Poster].

  • Searching to Exploit Memorization Effect in Learning from Noisy Labels.
    Q. Yao, H. Yang, B. Han, G. Niu, and J.T. Kwok.
    In Proceedings of 37th International Conference on Machine Learning (ICML'20), [PDF] [Code] [Poster].

  • Are Anchor Points Really Indispensable in Label-noise Learning?
    X. Xiao, T. Liu, N. Wang, B. Han, C. Gong, G. Niu, and M. Sugiyama.
    In Advances in Neural Information Processing Systems 32 (NeurIPS'19), [PDF] [Code] [Poster].

  • How does Disagreement Help Generalization against Label Corruption?
    X. Yu*, B. Han, J. Yao, G. Niu, I.W. Tsang, and M. Sugiyama.
    In Proceedings of 36th International Conference on Machine Learning (ICML'19), [PDF] [Code] [Poster] [Long Oral].

  • Efficient Nonconvex Regularized Tensor Completion with Structure-aware Proximal Iterations.
    Q. Yao, J.T. Kwok, and B. Han.
    In Proceedings of 36th International Conference on Machine Learning (ICML'19), [PDF] [Code] [Poster].

  • Towards Robust ResNet: A Small Step but A Giant Leap.
    J. Zhang*, B. Han, L. Wynter, B. Low, and M. Kankanhalli.
    In Proceedings of 28th International Joint Conference on Artificial Intelligence (IJCAI'19), [PDF] [Code] [Poster].

  • Privacy-preserving Stochastic Gradual Learning.
    B. Han, I.W. Tsang, X. Xiao, L. Chen, S.-F. Fung, and C. Yu.
    IEEE Transactions on Knowledge and Data Engineering (TKDE), 2019, [PDF].

  • Co-teaching: Robust Training of Deep Neural Networks with Extremely Noisy Labels.
    B. Han, Q. Yao, X. Yu, G. Niu, M. Xu, W. Hu, I.W. Tsang, and M. Sugiyama.
    In Advances in Neural Information Processing Systems 31 (NeurIPS'18), [PDF] [Code] [Poster] [Most Influential Paper Award].

  • Masking: A New Perspective of Noisy Supervision.
    B. Han, J. Yao, G. Niu, M. Zhou, I.W. Tsang, Y. Zhang, and M. Sugiyama.
    In Advances in Neural Information Processing Systems 31 (NeurIPS'18), [PDF] [Code] [Poster].

  • Millionaire: A Hint-guided Approach for Crowdsourcing.
    B. Han, Q. Yao, Y. Pan, I.W. Tsang, X. Xiao, Q. Yang, and M. Sugiyama.
    Machine Learning Journal (MLJ), 108(5): 831–858, 2018, [PDF] [Slides].

  • Stagewise Learning for Noisy k-ary Preferences.
    Y. Pan, B. Han, and I.W. Tsang.
    Machine Learning Journal (MLJ), 107(8): 1333–1361, 2018, [PDF].

  • Robust Plackett-Luce Model for k-ary Crowdsourced Preferences.
    B. Han, Y. Pan, and I.W. Tsang.
    Machine Learning Journal (MLJ), 107(4): 675–702, 2017, [PDF].


Brief Biography

    Bo Han is currently an Assistant Professor in Machine Learning and a Director of Trustworthy Machine Learning and Reasoning Group at Hong Kong Baptist University, and a BAIHO Visiting Scientist at RIKEN Center for Advanced Intelligence Project (RIKEN AIP), where his research focuses on machine learning, deep learning, and their applications. He was a Visiting Faculty Researcher at Microsoft Research (2022) and a Postdoc Fellow at RIKEN AIP (2019-2020). He received his Ph.D. degree in Computer Science from University of Technology Sydney (2015-2019). During 2018-2019, he was a Research Intern with the AI Residency Program at RIKEN AIP, working on trustworthy representation learning (e.g., Co-teaching and Masking). He also works on causal reasoning for trustworthy learning (e.g., CausalAdv and CausalNL). He has co-authored a machine learning monograph, including Machine Learning with Noisy Labels (MIT Press). He has served as Area Chairs of NeurIPS, ICML, ICLR and UAI. He has also served as Action (Associate) Editors of Transactions on Machine Learning Research and IEEE Transactions on Neural Networks and Learning Systems, and Editorial Board Members of Journal of Machine Learning Research and Machine Learning Journal. He received Outstanding Paper Award at NeurIPS and Outstanding Area Chair at ICLR. He received the RIKEN BAIHO Award (2019), RGC Early CAREER Scheme (2020), MSRA StarTrack Program (2021), and Tencent AI Faculty Research Award (2022).


Acknowledgement

    TMLR group is/was gratefully supported by UGC and RGC of Hong Kong, NSFC, GDST, RIKEN AIP, HKBU, HKBU CSD, CAAI, and industry research labs (Microsoft, Google, NVIDIA, Tencent, Alibaba).