AI RESEARCH
CARE: Class-Adaptive Expert Consensus for Reliable Learning with Long-Tailed Noisy Labels
arXiv CS.CV
•
ArXi:2605.23254v1 Announce Type: new Learning from real-world data is frequently hindered by the compound challenge of long-tailed class distributions and noisy annotations. Existing methods partially address these issues but typically ignore the non-uniform impact of label noise across classes, resulting in ineffective correction for tail classes and over-regularization for head classes.