Xiaoyan Zhang
I am a master's student in Electrical and Computer Engineering at the University of Michigan, advised by Qing Qu. My research interests focus on multi-agent systems, generative models, and large-scale representation learning. I received a B.E. degree in Artificial Intelligence from Anhui University. In Fall 2024, I collaborated with the Video and Image Processing Laboratory (VIPER) under the supervision of Fengqing Zhu, where I worked on continual learning and food recognition.
Email: xyzaxis AT umich.edu
Education
Selected Publications
SIGMA: Selective-Interleaved Generation with Multi-Attribute Tokens
Xiaoyan Zhang, Zechen Bai, Haofan Wang, Yiren Song
Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 2026
Paper / Code /
SIGMA introduces a selective-interleaved generation framework with multi-attribute tokens, enabling more controllable and compositional visual generation across diverse attributes.
Xiaoyan Zhang, Zechen Bai, Haofan Wang, Yiren Song
Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 2026
Paper / Code /
@article{zhang2026sigma,
title={SIGMA: Selective-Interleaved Generation with Multi-Attribute Tokens},
author={Zhang, Xiaoyan and Bai, Zechen and Wang, Haofan and Song, Yiren},
journal={arXiv preprint arXiv:2602.07564},
year={2026}
}SIGMA introduces a selective-interleaved generation framework with multi-attribute tokens, enabling more controllable and compositional visual generation across diverse attributes.
Dual-Imbalance Continual Learning for Real-World Food Recognition
Xiaoyan Zhang, Jiangpeng He
CVPR MetaFood Workshop, 2026
Best Paper Award
Paper / Code /
This paper studies dual-imbalance continual learning in real-world food recognition, addressing both class imbalance and data imbalance to improve robust incremental adaptation in practical scenarios.
Xiaoyan Zhang, Jiangpeng He
CVPR MetaFood Workshop, 2026
Best Paper Award
Paper / Code /
@article{zhang2026dual,
title={Dual-Imbalance Continual Learning for Real-World Food Recognition},
author={Zhang, Xiaoyan and He, Jiangpeng},
journal={arXiv preprint arXiv:2603.29133},
year={2026}
}This paper studies dual-imbalance continual learning in real-world food recognition, addressing both class imbalance and data imbalance to improve robust incremental adaptation in practical scenarios.
One Adapter for All: Towards Unified Representation in Step-Imbalanced Class-Incremental Learning
Xiaoyan Zhang, Jiangpeng He
arXiv preprint, 2026
Paper / Code /
One-A presents a unified adapter merging framework for step-imbalanced class-incremental learning, improving representation consistency while maintaining efficient deployment with a single merged adapter.
Xiaoyan Zhang, Jiangpeng He
arXiv preprint, 2026
Paper / Code /
@article{zhang2026one,
title={One Adapter for All: Towards Unified Representation in Step-Imbalanced Class-Incremental Learning},
author={Zhang, Xiaoyan and He, Jiangpeng},
journal={arXiv preprint arXiv:2603.10237},
year={2026}
}One-A presents a unified adapter merging framework for step-imbalanced class-incremental learning, improving representation consistency while maintaining efficient deployment with a single merged adapter.
Teaching Experience
- Fall 2024: ZX52340 Java Technology and Its Application (Practice), Teaching Assistant Anhui University
- Fall 2024: ZJ52014 Introduction to Artificial Intelligence, Teaching Assistant Anhui University
Awards
- Best Paper Award, CVPR 3rd MetaFood Workshop 2026
- Outstanding Graduate of Anhui Province, Ministry of Education of Anhui Province 2025
- National Scholarship, Ministry of Education of the People's Republic of China 2024
- Song Qingling Future Scholarship (0.05%), The China Song Qingling Foundation 2024
- Excellent Student Scholarship, Anhui University 2023
- Academic Science and Technology Scholarship, Anhui University 2023