Selective Posterior Margin Regularization for Forward-Corrected Classification
Learning with class-conditional label noise often relies on a transition model from latent clean classes to observed annotations. Forward correction embeds this transition in the likelihood, yet finite-sample networks may still memorize corrupted labels. The corrected likelihood also induces a reverse posterior over th...
Ze-Xing Zhang, Ji-Chao Li, Tian-Yang Lei et al.
· 0 citations