[Patent] Method and Apparatus for Federated Learning Using Loss-Based Relaxed Contrastive Learning with a Divergence Penalty on Same-Class Feature Vectors
Date 2025.01.07
Hit 117
Title : Method and Apparatus for Federated Learning Using Loss-Based Relaxed Contrastive Learning with a Divergence Penalty on Same-Class Feature Vectors
Application No. : PCT/KR2025/000282
Filing date : 2025년 01월 07일
Applicants : GCGenome; Seoul National University R&DB Foundation
Inventors : Bohyung Han, Sungwook Seo, Jingyu Kim, Jiho Kim
