BLOSSOM uses block-wise aggregation in multimodal federated learning to improve performance under sparse and missing modality conditions by up to 37.7%.
Compre- hensive semi-supervised multi-modal learning,
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.LG 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
BLOSSOM: Block-wise Federated Learning Over Shared and Sparse Observed Modalities
BLOSSOM uses block-wise aggregation in multimodal federated learning to improve performance under sparse and missing modality conditions by up to 37.7%.