Zedong Wang

Zedong Wang

Visiting Student at Westlake University

Westlake University

About

Please Visit My New Homepage For The Latest Updates!
I am an HK-born AI researcher. I obtained my B.Eng. in Electronic and Information Engineering (Machine Intelligence) from Huazhong University of Science & Technology. Currently, I am a visiting student in CAIRI AI Lab under Chair Prof. Stan Z. Li (IEEE Fellow, IAPR Fellow) at Westlake University. My research interests center around computer vision from 3 levels: (i) Data-level: Mixup augmentation & label efficient learning. (ii) Network-level: Efficient network architecture design. (iii) Framework-level: Vector Quantized representation learning and generation. Besides, I am also interested in AI for Genomics applications.

Previously, I worked on few-shot semantic segmentation fortunately supervised by Prof. Xinggang Wang. In the summer of 2021, I was a visiting student at Multimedia Lab (MMLab), Shenzhen Institute of Advanced Technology (SIAT), Chinese Academy of Sciences (CAS). Before that, I conducted a research internship remotely at Key Lab of Digital Earth Science, CAS (2020-2021).{.justify}

Welcome to contact me about research and internship.

News:

  • 04/2024 Please Visit My New Homepage For The Latest Updates!
  • 10/2023 New preprint SemiReward: The first reward model for general-purpose semi-supervised learning.
  • 09/2023 One paper OpenSTL is accepted by NeurIPS 2023.
  • 06/2023 Honored to be CSIG Student Member.
  • 05/2023 New preprint SAMix: The first mixup algorithm that solves the 2 remaining challenges at once for both SL and SSL scenarios.
  • 11/2022 New preprint MogaNet: A new family of pure ConvNet architecture (scaling from 5M to 100M+) explored from the novel perspective of multi-order game-theoretic interaction.
  • 09/2022 New preprint OpenMixup: The first comprehensive mixup benchmark and codebase framework for visual classification and more.
  • 09/2022 Responsible for the maintenance of OpenMixup codebase (525 stars).
Interests
  • Representation Learning
  • AI for Genomics
Education
  • B.Eng. in Electronic and Information Engineering, 2023

    Huazhong University of Science and Technology

Research Experience

 
 
 
 
 
Incoming Ph.D. Student
Jul 2022 – Present Hangzhou, Zhejiang, China

Responsibilities include:

  • Data mixing augmentation and label efficient leanring (OpenMixup, SAMix, SemiReward).
  • Efficient deep visual network architecture design (MogaNet).
  • Framework-level representation learning (OpenSTL).
  • AI for Genomics.
 
 
 
 
 
Research Intern
Sep 2021 – Jun 2022 Wuhan, Hubei, China

Responsibilities include:

  • Few-shot semantic segmentation.
 
 
 
 
 
Research Intern
Jul 2021 – Sep 2021 Shenzhen, Guangdong, China

Responsibilities include:

  • Semantic segmentation.
  • Text spotting
  • Vision model configuration.
  • Visual data sifting.
 
 
 
 
 
Research Intern (remote)
Sep 2020 – Apr 2021 Beijing, China

Responsibilities include:

  • High resolution remote sensing building semantic segmentation.

Projects (coming soon)

Deep learning projects with code

Recent Publications

Quickly discover relevant content by filtering publications.
(2022). Efficient Multi-order Gated Aggregation Network. Arxiv.

DOI

(2022). OpenMixup: Open Mixup Toolbox and Benchmark for Visual Representation Learning. Arxiv.

DOI

Contact