A skewness-guided pruning method removes skin-tone-related components in skin lesion classifiers, improving fairness and reducing computational cost.
• However, EOpp1 and EOdd showed significant group dif- ferences of approximately 7.4% and 8.2%, respectively, reflecting substantial fairness concerns
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
-
Enhancing Fairness in Skin Lesion Classification for Medical Diagnosis Using Prune Learning
A skewness-guided pruning method removes skin-tone-related components in skin lesion classifiers, improving fairness and reducing computational cost.