Privacy-Preserving Split Learning with Vision Transformers using Patch-Wise Random and Noisy CutMix
Privacy-Preserving Split Learning with Vision Transformers using Patch-Wise Random and Noisy CutMix
About this item
Full title
Author / Creator
Publisher
Ithaca: Cornell University Library, arXiv.org
Journal title
Language
English
Formats
Publication information
Publisher
Ithaca: Cornell University Library, arXiv.org
Subjects
More information
Scope and Contents
Contents
In computer vision, the vision transformer (ViT) has increasingly superseded the convolutional neural network (CNN) for improved accuracy and robustness. However, ViT's large model sizes and high sample complexity make it difficult to train on resource-constrained edge devices. Split learning (SL) emerges as a viable solution, leveraging server-sid...
Alternative Titles
Full title
Privacy-Preserving Split Learning with Vision Transformers using Patch-Wise Random and Noisy CutMix
Authors, Artists and Contributors
Identifiers
Primary Identifiers
Record Identifier
TN_cdi_proquest_journals_3088983731
Permalink
https://devfeature-collection.sl.nsw.gov.au/record/TN_cdi_proquest_journals_3088983731
Other Identifiers
E-ISSN
2331-8422
