Class-incremental Learning

Class-Incremental Learning is a continual learning paradigm in which a model incrementally acquires knowledge by sequentially learning new classes over time while maintaining the ability to recognize previously learned classes, with the distinguishing characteristic that the set of classes expands incrementally across tasks and the model must mitigate catastrophic forgetting under the constraint of limited or no access to prior class data.

231

Publications

16.6K

Citations

768

Authors

263

Institutions

Publications per year

2017–2026

224

Authors

768

  1. DZ
    Da-Wei Zhou

    Nanjing University, China

    12 publications · 1.2K citations

    12

    h-index

  2. JV
    Joost van de Weijer

    Universitat Autònoma de Barcelona, Spain

    12 publications · 1.3K citations

    11

    h-index

  3. HY
    Han-Jia Ye

    Nanjing University, China

    10 publications · 1.1K citations

    10

    h-index

  4. DZ
    De‐Chuan Zhan

    Nanjing University, China

    9 publications · 888 citations

    9

    h-index

  5. XL
    Xialei Liu

    Fifth Affiliated Hospital of Sun Yat-sen University, China

    9 publications · 1.3K citations

    9

    h-index

1–5 of 768

Venues

  1. arXiv (Cornell University)

    Repository · 39 publications · 1.9K citations

    19

    h-index

  2. 13

    h-index

  3. Proceedings of the AAAI Conference on Artificial Intelligence

    Conference · 11 publications · 608 citations

    11

    h-index

  4. IEEE Transactions on Pattern Analysis and Machine Intelligence

    TPAMI · 11 publications · 1.3K citations

    11

    h-index

  5. 7

    h-index

1–5