Marginal loss is most robust for multi-structure echo segmentation under inter-domain partial labels; aBCE is competitive for single missing labels; aCCE needs label dropout.
Title resolution pending
1 Pith paper cite this work, alongside 19 external citations. Polarity classification is still indexing.
1
Pith paper citing it
19
external citations · external index
fields
cs.CV 1years
2026 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Comparison of Loss Functions for Robust Deep Learning-based Echocardiography Segmentation when Learning with Partially Labelled Data from Multiple Domains
Marginal loss is most robust for multi-structure echo segmentation under inter-domain partial labels; aBCE is competitive for single missing labels; aCCE needs label dropout.