Split Learning

Split Learning is a distributed machine learning paradigm in which the forward and backward passes of neural network training are partitioned across devices, with the key differentiating characteristic that only a subset of layers is locally executed on client devices while the remaining layers reside on a server, thus distinguishing it from other federated or collaborative learning approaches by explicitly splitting the model architecture rather than keeping full or partial local models.

94

Publications

5.4K

Citations

327

Authors

146

Institutions

Publications per year

2017–2026

90

Authors

327

  1. PV
    Praneeth Vepakomma

    Motorola (United States), United States

    8 publications · 774 citations

    8

    h-index

  2. RR
    Ramesh Raskar

    Massachusetts Institute of Technology, United States

    8 publications · 774 citations

    8

    h-index

  3. CT
    Chandra Thapa

    Commonwealth Scientific and Industrial Research Organisation, Australia

    7 publications · 1K citations

    7

    h-index

  4. SC
    Seyit Camtepe

    Commonwealth Scientific and Industrial Research Organisation, Australia

    6 publications · 810 citations

    6

    h-index

  5. KH
    Kaibin Huang

    University of Hong Kong, Hong Kong SAR China

    4 publications · 280 citations

    4

    h-index

1–5 of 327

Venues

  1. arXiv (Cornell University)

    Repository · 20 publications · 1.1K citations

    14

    h-index

  2. IEEE Internet of Things Journal

    4 publications · 194 citations

    4

    h-index

  3. 3

    h-index

  4. 3

    h-index

  5. IEEE Wireless Communications

    2 publications · 133 citations

    2

    h-index

1–5