Currently, I am a third-year graduate student at the Shanghai Key Laboratory of Intelligent Sensing and Recognition, Shanghai Jiao Tong University(SJTU), advised by professor Tao Zhang. Before that, Iโve obtained the Bachelorโs degree in Electronic Engineering from Tianjin University.
I have some research experiences in optical/SAR/medical image processing, invariance/equivariance within neural networks as well as change detection, while currently I am interested in computer vision, specifically scene understanding and theoretical foundation.
Please feel free to contact me at xinyu.xu@sjtu.edu.cn for any communications related to my researches or anything else that I can help with.
๐ฅ News
- 2025.01๏ผย ๐๐ After 425 days, the article has finally been accepted by TIP!!! โจ
- 2024.11๏ผย ๐๐ Feel happy for receiving National Scholarship! ๐
- 2024.10๏ผย ๐๐ Feel happy for being honored as The Merit Student of Shanghai Jiao Tong University! ๐
- 2023.12๏ผย ๐๐ It was my honor to receive first-class academic scholarship from Shanghai Jiao Tong University. ๐ฎ
- 2023.11๏ผย ๐๐ It was my honor to receive the COSCO Shipping Scholarship from Shanghai Jiao Tong University. ๐ฎ
- 2022.09: ย ๐๐ Admitted by the School of Electronic Information and Electrical Engineering(SEIEE) of Shanghai Jiao Tong University! โจ
- 2022.06: ย ๐๐ Feel happy for graduation from TJU! ๐
๐ Educations
- 2022.09-2025.03: ย Master student in SJTU. Major: School of Electronic Information and Electrical Engineering
- 2018.09 -2022.06: ย Undergraduate student in TJU. Major: Electrical Engineering
๐ Publications
![sym](images/TIP.png)
PreCM: The Padding-based Rotation Equivariant Convolution Mode for Semantic Segmentation
Xinyu Xu, Huazhen Liu, Huilin Xiong, Wenxian Yu, Tao Zhang*.
IEEE Transactions on Image Processing, 2025.01
- Universal rotation equivariant convolution-group framework
- Component for replacing convolution
![sym](images/TCSVT.png)
Rotation Perturbation Robustness in Point Cloud Analysis: A Perspective of Manifold Distillation
Xinyu Xu, Huazhen Liu, Feiming Wei, Huilin Xiong, Wenxian Yu, Tao Zhang*.
IEEE Transactions on Circuits and Systems for Video Technology, 2024.08(In Peer Review)
- Online distillation
- Robustness against the rotation perturbation
![sym](images/IJS.png)
Development and validation of a novel deep ensemble learning-based fully automatic multi-structure segmentation framework for craniomaxillofacial surgery
Jiahao Baoโ , Zongcai Tanโ , Yifeng Sunโ , Xinyu Xuโ , et al.
International Journal of Surgery, 2024.09(In Peer Review)
- Segmentation of craniomaxillofacial structures
- Ensemble learning
![sym](images/IGARSS_2023.png)
An Information-expanding Network for Water Body Extraction based on U-net
Xinyu Xu, Huazhen Liu, Weiwei Guo, Zenghui Zhang, Tao Zhang*.
IEEE Geoscience and Remote Sensing Letters, 2023
- Rotation equivariant convolution
- Rotation-based channel attention mechanism
![sym](images/IGARSS_2024.png)
Flood Change Detection Based on Prior Feature Estimation
Xinyu Xu, Mingkang Xiong, Sinong Quan, et al.
IEEE International Geoscience and Remote Sensing Symposium, 2024
- Heterogeneous image
- Change detection
![sym](images/PIERS.png)
PolSAR Ship Detection with the Information Reconstruction-based Polarimetric Covariance Matrix
Xinyu Xu, Tao Zhang, Zenghui Zhang, Weiwei Guo, and Wenxian Yu.
PhotonIcs and Electromagnetics Research Symposium, 2023
- PolSAR ship detection
- Yamaguchi 4-decomposition
![sym](images/GRSL.png)
Water Body Detection Based on an Improved U-Net
Xinyu Xuโ , Huazhen Liuโ , Zenghui Zhang, Weiwei Guo, Tao Zhang*
IEEE International Geoscience and Remote Sensing Symposium, 2023
- Water body Segmentation
- Rotation equivariance
![Image 1](images/TJU_logo.jpg)
![Image 2](images/SJTU_logo.png)