Weakly Supervised Segmentation

Weakly Supervised Segmentation is an image segmentation paradigm that belongs to the broader class of supervised learning methods but is distinguished by its reliance on limited or inexact forms of annotation—such as image-level labels, bounding boxes, or scribbles—rather than requiring dense pixel-level ground truth masks for training.

610

Publications

52.7K

Citations

2.2K

Authors

608

Institutions

Publications per year

2017–2026

553

Authors

2.2K

  1. YW
    Yunchao Wei

    Beijing Jiaotong University, China

    16 publications · 3.3K citations

    14

    h-index

  2. XW
    Xinggang Wang

    Huazhong University of Science and Technology, China

    11 publications · 1.6K citations

    11

    h-index

  3. GL
    Guosheng Lin

    Nanyang Technological University, Singapore

    11 publications · 600 citations

    10

    h-index

  4. ZZ
    Zhaoxiang Zhang

    Chinese Academy of Sciences, China

    10 publications · 756 citations

    10

    h-index

  5. IB
    Ismail Ben Ayed

    École de Technologie Supérieure, Canada

    10 publications · 589 citations

    10

    h-index

1–5 of 2.2K

Venues

  1. arXiv (Cornell University)

    Repository · 96 publications · 8.9K citations

    30

    h-index

  2. 20

    h-index

  3. IEEE Transactions on Pattern Analysis and Machine Intelligence

    TPAMI · 23 publications · 1.9K citations

    19

    h-index

  4. IEEE Transactions on Medical Imaging

    20 publications · 2.1K citations

    18

    h-index

  5. Medical Image Analysis

    18 publications · 1.5K citations

    17

    h-index

1–5