Zeyue Xue

Zeyue Xue

AI advisor at Midjourney (China)

Biography

Hello, I’m Zeyue Xue, a researcher with a passion for building Generative AI products and platforms. I am currently an AI advisor at Midjourney (China), optimizing the Chinese version of Midjourney. Besides, I am also a PhD student at The University of Hong Kong. Previously, I spent two years at SenseTime Research (working on training large vision models), and used to be a research assistant at NUS (working on large-batch optimization).

We will announce some interesting projects in the near future. And my last representative work includes SenseMirage, a large-scale image generator introduced in RAPHAEL, which is the official text-to-image product by SenseTime.

My research interests lie in large-scale deep learning and deep generative models. If you are interested in my research, please feel free to email me.

Research Interests

  • Large-scale deep learning
  • Deep generative models

Publications

  1. Zeyue Xue*, Guanglu Song*, Qiushan Guo, Boxiao Liu, Zhuofan Zong, Yu Liu, Ping Luo. “RAPHAEL: Text-to-Image Generation via Large Mixture of Diffusion Paths”, NeurIPS 2023.
  2. Zeyue Xue, Jianming Liang, Guanglu Song, Zhuofan Zong, Liang Chen, Yu Liu, Ping Luo. “Large-batch Optimization for Dense Visual Predictions: Training Faster R-CNN in 4.2 minutes”, NeurIPS 2022.
  3. Zeyue Xue, Pan Zhou, Zichuan Xu, Xiumin Wang, Yulai Xie, Xiaofeng Ding, Shiping Wen. “A Resource-Constrained and Privacy-Preserving Edge-Computing-Enabled Clinical Decision System: A Federated Reinforcement Learning Approach”, IEEE IoT-J, 2021.
  4. Lingting Zhu, Zeyue Xue, Zhenchao Jin, Xian Liu, Jingzhen He, Ziwei Liu, Lequan Yu. “Make-a-Volume: Leveraging Latent Diffusion Models for Cross-Modality 3D Brain MRI Synthesis”, MICCAI 2023.
  5. Dazhong Shen, Guanglu Song, Zeyue Xue, Fu-Yun Wang, Yu Liu. “Rethinking the Spatial Inconsistency in Classifier-Free Diffusion Guidance”, CVPR 2024.
  6. Zhuofan Zong, Dongzhi Jiang, Guanglu Song, Zeyue Xue, Jingyong Su, Hongsheng Li, Yu Liu. “Temporal Enhanced Training of Multi-View 3D Object Detector via Historical Object Prediction”, ICCV 2023.