In federated learning with very few labels, a collaboratively trained diffusion model can synthesize balanced examples for missing classes and lift accuracy by about 14 points on CIFAR-10.
Convergence and accuracy trade-offs in federated learning and meta-learning,
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
citation-role summary
background 1
citation-polarity summary
fields
cs.LG 1years
2025 1verdicts
CONDITIONAL 1roles
background 1polarities
unclear 1representative citing papers
citing papers explorer
-
Diffusion Model-Based Data Synthesis Aided Federated Semi-Supervised Learning
In federated learning with very few labels, a collaboratively trained diffusion model can synthesize balanced examples for missing classes and lift accuracy by about 14 points on CIFAR-10.