AI RESEARCH
Redefining non-IID Data in Federated Learning for Computer Vision Tasks: Migrating from Labels to Embeddings for Task-Specific Data Distributions
arXiv CS.LG
•
ArXi:2503.14553v5 Announce Type: replace-cross Federated Learning (FL) has emerged as one of the prominent paradigms for distributed machine learning (ML). However, it is well-established that its performance can degrade significantly under non-IID (non-independent and identically distributed) data distributions across clients. To study this effect, the existing works predominantly emulate data heterogeneity by imposing label distribution skew across clients.