logo
CanyamAI summaries for academic research
language
MobileMobile
  • Phone FeatureWeChat Mini Program
EnglishEnglish
noticeNotice
BlogBlog
delete
arrow
Return

Federated unlearning using diffusive noise injection

delete2025-10-02
delete0
PRE
AI
M
Muhammad Usmani
M
Muhammad Atif Tahir
H
Humna Faisal
M
Muhammad Rafi
DOI：10.1016/j.inffus.2025.103796delete
deleteOriginal
deleteOriginal request for help
deleteShare
deleteSave
Abstract

Abstract

En 中文
• Removes client-level data in federated learning without central access. • Introduces Diffusive Noise Injection for unlearning and healing the global model after the unlearning process at the client level. • Maintains model accuracy while effectively unlearning specific contributions. • Tested on CIFAR, MNIST, and KVASIR using CNNs and Vision Transformers.

Journal

Information Fusion cover
Information Fusion
IF:
15.5
Papers:
4.3K
Citations:
2.7W

Organization

I
Institute of Business Administration
Scholars:
27
Papers: 25
Citations: 0
Cited Papers

Cited Papers

No cited papers available