Preprints

• How Does Learning Rate Decay Help Modern Neural Networks? [arxiv]
  • Kaichao You, Mingsheng Long, Jianmin Wang, Michael I. Jordan
• depyf: Open the Opaque Box of PyTorch Compiler for Machine Learning Researchers [arxiv] [Github]
  • Kaichao You, Runsheng Bai, Meng Cao, Jianmin Wang, Ion Stoica, Mingsheng Long
  • A PyTorch ecosystem project (see the introduction in https://pytorch.org/blog/introducing-depyf/ and PyTorch official twitter )
  • Highlighted in the opening keynote of PyTorch conference 2024

Journal Articles

• From Big to Small: Adaptive Learning to Partial-Set Domains [TPAMI] [arxiv] [IEEE Xplore]

  • Kaichao You*, Zhangjie Cao, Ziyang Zhang, Jianmin Wang, Mingsheng Long

• Ranking and Tuning Pre-trained Models: A New Paradigm for Exploiting Model Hubs [JMLR] [paper] [code]

  • Kaichao You, Yong Liu, Ziyang Zhang, Jianmin Wang, Michael I. Jordan, Mingsheng Long

• Tianshou: A Highly Modularized Deep Reinforcement Learning Library [JMLR] [arxiv] [code]

  • Jiayi Weng, Huayu Chen, Dong Yan, Kaichao You, Alexis Duburcq, Minghao Zhang, Hang Su, Jun Zhu
  • Github
  • Core developers

• Event-based Semantic Segmentation with Posterior Attention [TIP] [IEEE Xplore]

  • Zexi Jia, Kaichao You*, Weihua He, Yang Tian, Yongxiang Feng, Yaoyuan Wang, Xu Jia, Yihang Lou, Jingyi Zhang, Guoqi Li, Ziyang Zhang

Conference Proceedings

• Universal Domain Adaptation [CVPR2019] [pdf] [code] [poster]

  • Kaichao You, Mingsheng Long, Zhangjie Cao, Jianmin Wang, Michael I. Jordan

• Learning to Transfer Examples for Partial Domain Adaptation [CVPR2019] [pdf] [code] [poster]

  • Kaichao You*, Zhangjie Cao, Mingsheng Long, Jianmin Wang, Qiang Yang

• Towards Accurate Model Selection in Deep Unsupervised Domain Adaptation [ICML2019] [pdf] [code] [poster] [slides]

  • Kaichao You, Ximei Wang, Mingsheng Long, Michael I. Jordan

• Stochastic Normalization [NeurIPS2020] [code] [pdf]

  • Kaichao You*, Zhi Kou, Mingsheng Long, Jianmin Wang

• Co-Tuning for Transfer Learning [NeurIPS2020] [code] [pdf] [poster]

  • Kaichao You, Zhi Kou, Mingsheng Long, Jianmin Wang
  • Second Prize of Spectra Review Paper Competition, see the blog

• LogME: Practical Assessment of Pre-trained Models for Transfer Learning [ICML2021] [pdf] [code] [talk] [poster]

  • Kaichao You, Yong Liu, Jianmin Wang, Mingsheng Long
  • Included in Google Research.

• TimeReplayer: Unlocking the Potential of Event Cameras for Video Interpolation [CVPR2022] [arxiv]

  • Weihua He, Kaichao You, Zhendong Qiao, Xu Jia, Ziyang Zhang, Wenhui Wang, Huchuan Lu, Yaoyuan Wang, Jianxing Liao
  • Selected by Prophesee as a representative method in event-based slow motion. [Blog] [Image]

• Video Interpolation by Event-driven Anisotropic Adjustment of Optical Flow [ECCV2022] [arxiv]

  • Song Wu, Kaichao You*, Weihua He, Chen Yang, Yang Tian, Yaoyuan Wang, Jianxing Liao, Ziyang Zhang

• Test-Time Training-Free Domain Adaptation [ICASSP 2023] [IEEE Xplore]

  • Yongxiang Feng, Weihua He, Kaichao You, Bing Liu, Ziyang Zhang, Yaoyuan Wang, Minglei Li, Yihang Lou, Jiawei Li, Guoqi Li, Jianxing Liao

• Efficient ConvBN Blocks for Transfer Learning and Beyond [ICLR 2024] [arxiv] [code]

  • Kaichao You, Guo Qin, Anchang Bao, Meng Cao, Ping Huang, Jiulong Shan, Mingsheng Long
  • This feature has been already integrated into MMCV and PyTorch. It is available starting from PyTorch 2.2.

Professional Activities

• PC Member | Reviewer
  • TNNLS / WACV / IJCAI / AAAI / IEEE TIP / Neurocomputing / IEEE TPAMI / ICCV / CVPR / ICLR / ICML / NeurIPS

Invited Talks

• Present oral talks and posters at MLA 2019, November 2nd - 3rd 2019

  • [slides]

• Advances in Domain Adaptation: Setting, Method and Model Selection at BAAI Live, Dec. 24, 2019

  • [slides]
  • [Video Link]

• LogME: Practical Assessment of Pre-trained Models for Transfer Learning at BAAI Live, Mar. 10, 2021

  • [slides]

• Efficient Transfer Learning Across Data, Model, and Modality at Apple APAC AI/ML tech talk. Mar. 9, 2023

Competition

• VisDA2018 Classification Track, 3rd place [certificate] [poster] [slides]``