Weiming Zhuang

Weiming Zhuang

Research Scientist

Sony AI

Biography

I am a research scientist at Sony AI, focusing on vision foundation model and federated learning in Privacy-Preserving Machine Learning (PPML) team.

Before joining Sony AI, I was a Ph.D. researcher under SenseTime-NTU Talent Programme and obtained Ph.D. from Nanyang Technological University, advised by Prof. Yonggang Wen. Prior to that, I had years of experience in software engineering, focusing on build large-scale distributed systems. I obtained my Bachelor’s from the School of Computing, National University of Singapore.

Interests
  • Vision Foundation Model
  • Federated Learning
  • Computer Vision
  • Machine Learning System
  • Efficient On-device ML
Education
  • Ph.D., 2019 - 2022

    Nanyang Technological University

  • BSc(Hons) in Information System, 2013 - 2016

    National University of Singapore

News

  • [2024-07]: One paper is accepted by ECCV'24. This work introduces a new data synthesis method to improve object detection by augmenting image backgrounds.
  • [2024-04]: One paper is accepted by ICML'24. This work introduces a new practical and vision-centric federated learning platform
  • [2024-04]: One paper is accepted by ICS'24. This work is about scheduling foundation model fine-tuning workloads in datacenters.
  • [2024-01]: One paper is accepted by CVPR'24. This work is about memory-efficient federated dynamic pruning.
  • [2024-01]: One paper is accepted by ICLR'24. This work is about federated learning without normalization for feature shift problem.
  • [2023-12]: One paper is accepted by AAAI'24. This work is about federated semi-supervised learning.
  • [2023-09]: One paper is accepted by NeurIPS'23. This work is about handling test-time shift in federated learning.
  • [2023-07]: Two papers have been accepted to ICCV'23. They are on federated multiple-task learning and federate continual learning.
  • [2023-07]: One paper is accepted by KDD-FL4DataMining'23 (Best Industry Paper Award) and FL-IJCAI'23. This work is about federated learning without normalizations.
  • [2022-11]: I joined Sony AI as a Research Scientist.
  • [2022-02]: One paper is accepted by ICLR'22. This work is about federated self-supervised learning.
  • [2022-01]: One paper is published by IEEE Internet of Things Journal. This work is about federated learning platform.
  • [2021-07]: One paper is accepted by ICCV'21. This work is about federated self-supervised learning.
  • [2021-07]: One paper is accepted by ACMMM'21. This work is about federated unsupervised person reid.