Conference paper2025ICC 2025 - IEEE International Conference on Communications

ProFe: Communication-Efficient Decentralized Federated Learning via Distillation and Prototypes

TrainingQuantization (signal)CostsFederated learningPrototypesCollaborationData modelsComplexity theoryOptimizationFacesCommunication OptimizationFederated LearningKnowledge DistillationPrototype LearningQuantization

Quick facts

Year
2025
Venue
ICC 2025 - IEEE International Conference on Communications
Identifier
sanchezsanchez2025profe

Suggested citation

Pedro Miguel Sánchez Sánchez, Enrique Tomás Martínez Beltrán, Miguel Fernández Llamas, Gérôme Bovet, Gregorio Martínez Pérez, Alberto Huertas Celdrán (2025). ProFe: Communication-Efficient Decentralized Federated Learning via Distillation and Prototypes. ICC 2025 - IEEE International Conference on Communications.

Abstract

Authors

Pedro Miguel Sánchez SánchezEnrique Tomás Martínez BeltránMiguel Fernández LlamasGérôme BovetGregorio Martínez PérezAlberto Huertas Celdrán

Keywords

TrainingQuantization (signal)CostsFederated learningPrototypesCollaborationData modelsComplexity theoryOptimizationFacesCommunication OptimizationFederated LearningKnowledge DistillationPrototype LearningQuantization

Related publications

Works with stronger overlap in topic, type, and tags.

Related Research