Jie Zhu

Haidian district, Beijing

Jie Zhu currently is a fifth-year Ph.D. candidate in the School of Computer Science, Peking University, supervised by tenured Associate Professor Leye Wang. Prior to that, he obtained his bachelor degree from Beihang University in 2021.

His research interests mainly involve multi-modality large language model (MLLM), e.g., visual perception, generative models, and their unification, and AI security, e.g., membership inference and data privacy.

He previously interned at Megvii Technology (2020.12-2021.12), Baidu VIS (2022.7-2025.5), and Meituan(2025.5-2026.3) and has published six first-author papers (5 CCF-A) including CCS, ICLR, NeurIPS, ASE, IEEE TSE, and TMLR.

selected publications

NeurIPS

Mole: Enhancing human-centric text-to-image diffusion via mixture of low-rank experts

Jie Zhu, Yixiong Chen, Mingyu Ding, Ping Luo, Leye Wang, and Jingdong Wang

Advances in Neural Information Processing Systems, 2024

Bib PDF

@article{zhu2024mole,
  title = {Mole: Enhancing human-centric text-to-image diffusion via mixture of low-rank experts},
  author = {Zhu, Jie and Chen, Yixiong and Ding, Mingyu and Luo, Ping and Wang, Leye and Wang, Jingdong},
  journal = {Advances in Neural Information Processing Systems},
  volume = {37},
  pages = {29354--29386},
  year = {2024},
}

ACM CCS

A unified membership inference method for visual self-supervised encoder via part-aware capability

Jie Zhu, Jirong Zha, Ding Li, and Leye Wang

In Proceedings of the 2024 on ACM SIGSAC Conference on Computer and Communications Security, 2024

Bib PDF

@inproceedings{zhu2024unified,
  title = {A unified membership inference method for visual self-supervised encoder via part-aware capability},
  author = {Zhu, Jie and Zha, Jirong and Li, Ding and Wang, Leye},
  booktitle = {Proceedings of the 2024 on ACM SIGSAC Conference on Computer and Communications Security},
  pages = {1241--1255},
  year = {2024},
}

IEEE TSE

Safety and performance, why not both? bi-objective optimized model compression against heterogeneous attacks toward ai software deployment

Jie Zhu, Leye Wang, Xiao Han, Anmin Liu, and Tao Xie

IEEE Transactions on Software Engineering, 2024

Bib PDF

@article{zhu2024safety,
  title = {Safety and performance, why not both? bi-objective optimized model compression against heterogeneous attacks toward ai software deployment},
  author = {Zhu, Jie and Wang, Leye and Han, Xiao and Liu, Anmin and Xie, Tao},
  journal = {IEEE Transactions on Software Engineering},
  volume = {50},
  number = {3},
  pages = {376--390},
  year = {2024},
  publisher = {IEEE},
}

TMLR

Understanding Self-Supervised Pretraining with Part-Aware Representation Learning

Jie Zhu, Jiyang Qi, Mingyu Ding, Xiaokang Chen, Ping Luo, Xinggang Wang, and 3 more authors

Transactions on Machine Learning Research, 2023

Bib PDF

@article{zhuunderstanding,
  title = {Understanding Self-Supervised Pretraining with Part-Aware Representation Learning},
  author = {Zhu, Jie and Qi, Jiyang and Ding, Mingyu and Chen, Xiaokang and Luo, Ping and Wang, Xinggang and Liu, Wenyu and Wang, Leye and Wang, Jingdong},
  journal = {Transactions on Machine Learning Research},
  year = {2023},
}

ICLR

E-CRF: Embedded Conditional Random Field for Boundary-caused Class Weights Confusion in Semantic Segmentation

Jie Zhu, Huabin Huang, Banghuai Li, and Leye Wang

In The Eleventh International Conference on Learning Representations, 2023

Bib PDF

@inproceedings{zhucrf,
  title = {E-CRF: Embedded Conditional Random Field for Boundary-caused Class Weights Confusion in Semantic Segmentation},
  author = {Zhu, Jie and Huang, Huabin and Li, Banghuai and Wang, Leye},
  booktitle = {The Eleventh International Conference on Learning Representations},
  year = {2023},
}

ASE

Safety and performance, why not both? bi-objective optimized model compression toward ai software deployment

Jie Zhu, Leye Wang, and Xiao Han

In Proceedings of the 37th IEEE/ACM international conference on automated software engineering, 2022

Bib PDF

@inproceedings{zhu2022safety,
  title = {Safety and performance, why not both? bi-objective optimized model compression toward ai software deployment},
  author = {Zhu, Jie and Wang, Leye and Han, Xiao},
  booktitle = {Proceedings of the 37th IEEE/ACM international conference on automated software engineering},
  pages = {1--13},
  year = {2022},
}