Masked Image Modeling

Masked Image Modeling is a self-supervised learning framework that belongs to the class of pretext tasks in computer vision, distinguished by its strategy of reconstructing randomly masked image patches from the visible portions of an input image to learn expressive visual representations.

119

Publications

8.1K

Citations

554

Authors

138

Institutions

Publications per year

2017–2026

114

Authors

554

  1. HH
    Han Hu

    Microsoft Research Asia (China), China

    7 publications · 1.4K citations

    7

    h-index

  2. ZZ
    Zheng Zhang

    Harbin Institute of Technology, China

    6 publications · 1.4K citations

    6

    h-index

  3. YC
    Yue Cao

    Microsoft Research Asia (China), China

    4 publications · 1.3K citations

    4

    h-index

  4. ZX
    Zhenda Xie

    Tsinghua University, China

    4 publications · 1.3K citations

    4

    h-index

  5. DC
    Dongdong Chen

    University of Science and Technology of China, China

    4 publications · 393 citations

    4

    h-index

1–5 of 554

Venues

  1. arXiv (Cornell University)

    Repository · 34 publications · 2K citations

    17

    h-index

  2. 10

    h-index

  3. Proceedings of the AAAI Conference on Artificial Intelligence

    Conference · 4 publications · 149 citations

    4

    h-index

  4. Remote Sensing

    3 publications · 48 citations

    3

    h-index

  5. 3

    h-index

1–5