arXiv preprint 2019

Robust Aggregation for Federated Learning

arXiv preprint 2019 krishnap25/RFA

The proposed approach relies on a robust secure aggregation oracle based on the geometric median, which returns a robust aggregate using a constant number of calls to a regular non-robust secure average oracle.

Robust Aggregation for Federated Learning

arXiv preprint 2019 krishnap25/RFA

The proposed approach relies on a robust secure aggregation oracle based on the geometric median, which returns a robust aggregate using a constant number of calls to a regular non-robust secure average oracle.

Meta-Learning Initializations for Image Segmentation

arXiv preprint 2019 ml4ai/mliis

Our primary contributions include (1), an extension and experimental analysis of first-order model agnostic meta-learning algorithms (including FOMAML and Reptile) to image segmentation, (2) a novel neural network architecture built for parameter efficiency and fast learning which we call EfficientLab, (3) a formalization of the generalization error of meta-learning algorithms, which we leverage to decrease error on unseen tasks, and (4) a small benchmark dataset, FP-k, for the empirical study of how meta-learning systems perform in both few- and many-shot settings.

FEW-SHOT IMAGE CLASSIFICATION FEW-SHOT SEMANTIC SEGMENTATION SEMANTIC SEGMENTATION