LONGHUI YU

Biography

Longhui Yu is a MPhil student at Peking University, advised by Prof. Yuesheng Zhu. Currently, I am sincerely looking for a Ph.D. opportunity. Here is my CV. I am very lucky to work with Dr. Weiyang Liu from University of Cambridge & Max Planck Institute and Dr. Lanqing Hong from National University of Singapore.

My research interest broadly lies in Machine Learning methods and applications. Previously, I spent wonderful time with my collaborators working on Continual Learning, Long‑tailed Learning, Semi‑supervised Learning, Neural Collapse, Out‑of‑Distribution Detection. Meanwhile, I also pay attention to the real-world applications of deep learning, such as Autonomous Driving, Object Detection, NeRF, and Vector Font Synthesis.

For my Ph.D research, I am highly interested in Trustworthy AI, Human‑centric/Data‑centric AI, Interactive ML, AI for Healthcare, Implicit Representation. In addition, I am also interested in some hot techniques. For examples, How do we learn knowledge from large models? How do we propose some methods to replace SGD? How do we link generative models and classification models? How do LLMs help XAI and AGI? In what scientific fields does the current Deep Learning still have great potential?

Currently, I am sincerely looking for a Ph.D. opportunity!

Publication and Preprints

2022

• Longhui Yu, Tianyang Hu, Lanqing Hong, Zhen Liu, Adrian Weller, Weiyang Liu.
  Continual Learning by Modeling Intra‑Class Variation. [PDF]
  In Publications of Transactions on Machine Learning Research (TMLR), 2023.
  
• Weiyang Liu*, Longhui Yu*, Adrian Weller, Bernhard Schölkopf.
  Generalizing and Decoupling Neural Collapse via Hyperspherical Uniformity Gap. [PDF]
  In Proceedings of International Conference on Learning Representations (ICLR), 2023.
  
• Longhui Yu, Yifan Zhang, Lanqing Hong, Fei Chen, Zhenguo Li.
  Dual-Curriculum Teacher for Domain-Inconsistent Object Detection in Autonomous Driving. [PDF]
  In Proceedings of British Machine Vision Conference (BMVC), 2022.
  
• Longhui Yu, Zhenyu Weng, Yuqing Wang, Yuesheng Zhu.
  Multi-Teacher Knowledge Distillation for Incremental Implicitly-Refined Classification. [PDF]
  In Proceedings of International Conference on Multimedia and Expo (ICME Oral), 2022.
  
• Liyuan Wang*, Xingxing Zhang*, Kuo Yang, Longhui Yu, Chongxuan Li, Lanqing Hong, Shifeng Zhang, Zhenguo Li, Yi Zhong, Jun Zhu.
  Memory Replay with Data Compression for Continual Learning. [PDF]
  In Proceedings of International Conference on Learning Representations (ICLR), 2022.
  
• Shoukang Hu, Kaichen Zhou, Longhui Yu, Lanqing Hong, Tianyang Hu, Gim Hee Lee, Zhenguo Li.
  MaskNeRF: Masked Neural Radiance Fields for Sparse View Synthesis. [PDF]
  In Submissions of IEEE Conference on Computer Vision and Pattern Recognition (CVPR Submissions), 2023.
  
• Yuqing Wang, Yizhi Wang, Longhui Yu, Yuesheng Zhu, Zhouhui Lian.
  DeepVecFont-v2: Exploiting Transformers to Synthesize Vector Fonts with Higher Quality. [PDF]
  In Submissions of IEEE Conference on Computer Vision and Pattern Recognition (CVPR Submissions), 2023.
  

Awards

• 2020, Outstanding Graduate, The Outstanding Graduate of South China University of Technology
• 2019, The First Prize, The 14th National University Students’ Smart Car Competition
• 2019, Scholarship, The Scholarship of South China University of Technology
• 2018, Scholarship, The Scholarship of South China University of Technology
• 2017, Scholarship, The Scholarship of South China University of Technology

Professional Service

• Reviewer
  • NeurIPS 2023, Conference on Neural Information Processing Systems
  • ICCV 2023, IEEE International Conference on Computer Vision
  • CVPR 2023, IEEE Conference on Computer Vision and Pattern Recognition
  • ACML 2022, Asian Conference on Machine Learning
  • BMVC 2022, British Machine Vision Conference