A training framework that reweights samples by the gap between ground-truth and nearest-negative prediction scores, plus flip consistency, improves FER accuracy under label noise and class imbalance.
J.; Erhan, D.; Carrier, P
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CV 1years
2025 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Navigating Label Ambiguity for Facial Expression Recognition in the Wild
A training framework that reweights samples by the gap between ground-truth and nearest-negative prediction scores, plus flip consistency, improves FER accuracy under label noise and class imbalance.