A class-specific threshold learning method, driven by quantiles of the model's own score distributions plus a ranking loss, improves pseudo-label quality in multi-label recognition with partial labels.
Dy- namic correlation learning and regularization for multi-label confidence calibration
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
citation-role summary
background 1
citation-polarity summary
fields
cs.CV 1years
2025 1verdicts
CONDITIONAL 1roles
background 1polarities
unclear 1representative citing papers
citing papers explorer
-
Learning Semantic-Aware Threshold for Multi-Label Image Recognition with Partial Labels
A class-specific threshold learning method, driven by quantiles of the model's own score distributions plus a ranking loss, improves pseudo-label quality in multi-label recognition with partial labels.