Few-shot Defect Detection

Few-Shot Defect Detection is a defect detection paradigm that belongs to the class of machine learning approaches in which a model identifies anomalies or defects in new categories by learning from a very limited number of labeled defective instances, distinguishing itself through its reliance on minimal supervision compared to conventional defect detection methods requiring large annotated datasets.

31

Publications

1.5K

Citations

119

Authors

38

Institutions

Publications per year

2017–2026

30

Authors

119

  1. KS
    Kechen Song

    Northeastern University, China

    3 publications · 169 citations

    3

    h-index

  2. YY
    Yunhui Yan

    Northeastern University, China

    3 publications · 169 citations

    3

    h-index

  3. DK
    Dermot Kerr

    University of Ulster, United Kingdom

    2 publications · 50 citations

    2

    h-index

  4. GD
    Guifang Duan

    Zhejiang University, China

    2 publications · 56 citations

    2

    h-index

  5. JT
    Jianrong Tan

    Zhejiang University, China

    2 publications · 56 citations

    2

    h-index

1–5 of 119

Venues

  1. 4

    h-index

  2. IEEE Transactions on Industrial Informatics

    3 publications · 102 citations

    3

    h-index

  3. Journal of Intelligent Manufacturing

    2 publications · 58 citations

    2

    h-index

  4. Computer-Aided Civil and Infrastructure Engineering

    CACAIE · 2 publications · 160 citations

    2

    h-index

  5. Computers in Industry

    2 publications · 84 citations

    2

    h-index

1–5