Lavie’s Homepage

Biography

I am a researcher interested in mathematical and statistical methods for distribution shift and transfer learning, including learning theory, kernel methods, and optimal transport. I received my Ph.D. degree in Applied Math under the guidance of Prof. Chuan-Xian Ren from SYSU. Previously, I received my B.S. degree under the supervision of Prof. Gang Wu from CUMT, where I focused on the application of matrix theory.

Education

Ph.D. in Applied Math, 2018 - 2023
School of Mathematics, Sun Yat-Sen University, Guangzhou, China.
Prof. Chuan-Xian Ren

B.S. in Statistics, 2014 - 2018
School of Mathematics, China University of Mining and Technology, Xuzhou, China.
Prof. Gang Wu

Selected Publications

Dataset Shift and Generalization Analysis:

When Invariant Representation Learning Meets Label Shift: Insufficiency and Theoretical Insights
You-Wei Luo and Chuan-Xian Ren*.
IEEE Transactions on Pattern Analysis and Machine Intelligence (TPAMI). 2024.
[IEEE] [arXiv]
“A systematic study of invariant representation learning with GLS correction, where the theoretical sufficiency and necessity are provided.”
Geometric Understanding of Discriminability and Transferability for Visual Domain Adaptation
You-Wei Luo, Chuan-Xian Ren*, Xiao-Lin Xu and Qingshan Liu.
IEEE Transactions on Pattern Analysis and Machine Intelligence (TPAMI). 2024.
[IEEE] [arXiv]
“The co-regularization between discriminability and transferability, which ensures the existence of optimal representations with simultaneously maximized two abilities.”
Unsupervised Domain Adaptation via Discriminative Manifold Propagation
You-Wei Luo, Chuan-Xian Ren*, Dao-Qing Dai and Hong Yan.
IEEE Transactions on Pattern Analysis and Machine Intelligence (TPAMI). 2022, 44(3): 1653-1669.
[IEEE] [arXiv]
“We propose a unified manifold learning framework for the UDA and PDA problems, and prove the error bounds with the metrics on the different types of manifolds for both DA settings.”
Unsupervised Domain Adaptation via Discriminative Manifold Embedding and Alignment
You-Wei Luo, Chuan-Xian Ren*, Pengfei Ge, Ke-Kun Huang and Yu-Feng Yu.
Proceedings of the AAAI Conference on Artificial Intelligence (AAAI Oral). 2020.
[AAAI] [arXiv] [Slides] [Code]
“DRMEA describes the domains by a sequence of abstract manifolds, and develops a Riemannian manifold learning framework to achieve transferability and discriminability consistently.”

Optimal Transport:

MOT: Masked Optimal Transport for Partial Domain Adaptation
You-Wei Luo and Chuan-Xian Ren*.
Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR). 2023.
[CVF] [Preprint] [Video]
“A novel mechanism to overcome strict marginal constraints in OT and achieve conditional transfer.”
BuresNet: Conditional Bures Metric for Transferable Representation Learning
Chuan-Xian Ren*, You-Wei Luo and Dao-Qing Dai.
IEEE Transactions on Pattern Analysis and Machine Intelligence (TPAMI). 2023, 45(4): 4198-4213.
[IEEE]
“A plug-and-play discrepancy optimization module for transfer learning scenarios, e.g., domain adaptation and few-shot learning.”
Conditional Bures Metric for Domain Adaptation
You-Wei Luo and Chuan-Xian Ren*.
Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR). 2021.
[CVF] [Preprint] [Poster] [Video] [Code]
“We develop a theoretical conditional distribution discrepancy called Conditional Kernel Bures (CKB) metric, and propose a conditional invariant feature learning model for UDA.”
Towards Unsupervised Domain Adaptation via Domain-Transformer
Chuan-Xian Ren*, Yi-Ming Zhai, You-Wei Luo and Hong Yan.
International Journal of Computer Vision (IJCV) . 2024.
[Springer]
“We connect the core mechanism of Transformer with the optimal transport, where the generalization error can be controlled by the cross-domain Wasserstein distance.”

Academic Service

Reviewer or PC Member:

IEEE Transactions on Pattern Analysis and Machine Intelligence (IEEE TPAMI)
IEEE Transactions on Neural Networks and Learning Systems (IEEE TNNLS)
International Conference on Machine Learning (ICML): 2022 (Outstanding Top 10%), 2023, 2024
Advances in Neural Information Processing Systems (NeurIPS): 2022, 2023, 2024
International Conference on Learning Representations (ICLR): 2024, 2025
IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR): 2023, 2024
European Conference on Computer Vision (ECCV): 2024
International Conference on Artificial Intelligence and Statistics (AISTATS): 2025
AAAI Conference on Artificial Intelligence (AAAI): 2025

All Publications

[TPAMI’24b] You-Wei Luo, Chuan-Xian Ren*. When Invariant Representation Learning Meets Label Shift: Insufficiency and Theoretical Insights. IEEE TPAMI, 2024. [IEEE] [arXiv]
[TPAMI’24a] You-Wei Luo, Chuan-Xian Ren*, Xiao-Lin Xu, Qingshan Liu. Geometric Understanding of Discriminability and Transferability for Visual Domain Adaptation. IEEE TPAMI, 2024. [IEEE] [arXiv]
[IJCV’24] Chuan-Xian Ren*, Yi-Ming Zhai, You-Wei Luo, Hong Yan. Towards Unsupervised Domain Adaptation via Domain-Transformer. IJCV, 2024. [Springer]
[ECCV’24] Hao-Ran Yang, Chuan-Xian Ren*, You-Wei Luo. COD: Learning Conditional Invariant Representation for Domain Adaptation Regression. ECCV (Oral), 2024. [arXiv]
[AAAI’24] Yan Wang, Chuan-Xian Ren*, Yi-Ming Zhai, You-Wei Luo, Hong Yan. Probability-Polarized Optimal Transport for Unsupervised Domain Adaptation. AAAI, 2024. [AAAI]
[SCIS’24] Yi-Ming Zhai, Chuan-Xian Ren*, You-Wei Luo, Dao-Qing Dai. Maximizing Conditional Independence for Unsupervised Domain Adaptation. SCIS, 2024, 67(5): 152108. [Springer]
[CVPR’23] You-Wei Luo, Chuan-Xian Ren*. MOT: Masked Optimal Transport for Partial Domain Adaptation. CVPR, 2023. [CVF] [Preprint] [Video]
[TPAMI’23] Chuan-Xian Ren*, You-Wei Luo, Dao-Qing Dai. BuresNet: Conditional Bures Metric for Transferable Representation Learning. IEEE TPAMI, 2023, 45(4): 4198-4213. [IEEE]
[TPAMI’22] You-Wei Luo, Chuan-Xian Ren*, Dao-Qing Dai, Hong Yan. Unsupervised Domain Adaptation via Discriminative Manifold Propagation. IEEE TPAMI, 2022, 44(3): 1653-1669. [IEEE] [arXiv]
[CVPR’21] You-Wei Luo, Chuan-Xian Ren*. Conditional Bures Metric for Domain Adaptation. CVPR, 2021. [CVF] [Preprint] [Poster] [Video] [Code]
[AAAI’20] You-Wei Luo, Chuan-Xian Ren*, Pengfei Ge, Ke-Kun Huang, Yu-Feng Yu. Unsupervised Domain Adaptation via Discriminative Manifold Embedding and Alignment. AAAI (Oral), 2020. [AAAI] [arXiv] [Slides] [Code]
[TIP’20] Chuan-Xian Ren*, You-Wei Luo, Xiao-Lin Xu, Dao-Qing Dai, Hong Yan. Discriminative Residual Analysis for Image Set Classification With Posture and Age Variations. IEEE TIP, 2020, 29: 2875-2888. [IEEE] [arXiv] [Code]
[Calcolo’20] Wenya Shi, You-Wei Luo, Gang Wu*. On General Matrix Exponential Discriminant Analysis Methods for High Dimensionality Reduction. Calcolo, 2020, 57(2). [Springer] [Preprint]
[CVPR’20] Mengxue Li, Yi-Ming Zhai, You-Wei Luo, Peng-Fei Ge, Chuan-Xian Ren*. Enhanced Transport Distance for Unsupervised Domain Adaptation. CVPR, 2020. [CVF] [IEEE] [Poster] [Slides] [Code]