Under this setting, previous SSL methods tend to predict wrong pseudo-labels with the model fitted on labeled data, resulting in noise accumulation.

Previous approaches often downweight instances from irrelevant classes to mitigate the negative impact of class distribution mismatch on model training.

CARP learns prototypes in an end-to-end online fashion using gradient descent without additional non-differentiable modules to solve the cluster assignment problem. CARP optimizes a new pretext task based on random partitions of prototypes that regularizes the model and enforces consistency between views' assignments.

In doing so, we are able to find the exact settings for which different methodsprovablybecome identical.

Mostexisting results on Federated Learning (FL) assume the clients have ground-truth labels.