Xiawu Zheng （郑侠武）

About me

Xiawu Zheng is currently an associate professor in Xiamen University. He was a  postdoc researcher, working with YonghongTian, Rongrong Ji and Jie Chen in Pengcheng National Lab. He received Ph.D. (3.5-year early graduation) from  Xiamen University under the supervision of Prof. Rongrong Ji. His research interests include Model Compression, Neural Architecture Search, and Automated Machine Learning.

News

Publications

Journal

1. Xiawu Zheng, Rongrong Ji, Yuhang Chen, Qiang Wang, Baochang Zhang, Qixiang Ye, Jie Chen, Feiyue Huang, Yonghong Tian. MIGO-NAS: Towards Fast and Generalizable Neural Architecture Search. IEEE Transactions on Pattern Analysis and Machine Intelligence (T-PAMI) 2021.

2. Xiawu Zheng, Yang Zhang, Sirui Hong, Huixia Li, Lang Tang, Youcheng Xiong, Jin Zhou, Yan Wang, Xiaoshuai Sun, Pengfei Zhu, Chengling Wu, Rongrong Ji. Evolving Fully Automated Machine Learning via Life-Long Knowledge Anchors. IEEE Transactions on Pattern Analysis and Machine Intelligence (T-PAMI) 2021.

3. Xiawu Zheng, Chenyi Yang, Shaokun Zhang, Yan Wang, Baochang Zhang, Yongjian Wu, Yunsheng Wu, Ling Shao, Rongrong Ji. DDPNAS: Efficient Neural Architecture Search via Dynamic Distribution Pruning. International Journal of Computer Vision (IJCV) 2023.

4. Xiawu Zheng, Yuexiao Ma, Teng Xi, Gang Zhang, Errui Ding, Yuchao Li, Jie Chen, Yonghong Tian, Rongrong Ji∗. An Information Theory-inspired Strategy for Automatic Network Pruning. International Journal of Computer Vision (IJCV) 2025.

5. Xiawu Zheng, Lei Zhang, Binghan Chen, Mingkai Wang, Fei Chao, Chenglin Wu, Shanshan Wang, Rongrong Ji, Yonghong Tian. Good Performance Estimation Strategies Are All You Need in Neural Architecture Search. IEEE Transactions on Pattern Analysis and Machine Intelligence (T-PAMI) , 2025.

Academic Competitions

• AutoDL 2019: the final challenge in 2019 AutoDL challenges series, Rank 1st.

• AutoCV2 2019: IMAGE + VIDEO Automated Computer Vision challenge, Rank 2st / 34 teams.

 Projects

• MetaGPT , which assign different roles to GPTs to form a collaborative software entity for complex tasks.

• XNAS is an effective, modular and flexible neural architecture search (NAS) codebase, which aims to provide a common framework and baselines for the NAS community.

• XCompression is An easy use Neural Network Compression framework.

• XBBO is an effective, modular and flexible black-box optimization (BBO) codebase, which aims to provide a common framework and benchmark for the BBO community.