Idioma:

DEGAN: Decentralized generative adversarial networks

Faezi, Mohammad Hashem ; Bijani, Shahriar ; Dolati, Ardeshir

Neurocomputing (Amsterdam), 2021-01, Vol.419, p.335-343 [Periódico revisado por pares]

Texto completo disponível

Citações Citado por

Título:
DEGAN: Decentralized generative adversarial networks
Autor: Faezi, Mohammad Hashem ; Bijani, Shahriar ; Dolati, Ardeshir
Assuntos: Decentralized architecture ; Deep learning ; Distributed machine learning ; Generative adversarial networks
É parte de: Neurocomputing (Amsterdam), 2021-01, Vol.419, p.335-343
Descrição: We propose a distributed and decentralized Generative Adversarial Networks (GANs) framework without the exchange of the training data. Each node contains local dataset, a discriminator and a generator, from which only the generator gradients are shared with other nodes. In this paper, we introduce a novel, distributed technique in which workers communicate directly with each other, having no central nodes. Our experimental results on the benchmark datasets demonstrate almost the same performance and accuracy compared with existing centralized GAN frameworks. The proposed framework addresses the lack of decentralized learning for GANs.
Editor: Elsevier B.V
Idioma: Inglês

Voltar para lista de resultados