PulseAugur
EN
LIVE 11:46:51
research · [2 sources] ·

New CARE framework improves AI learning with noisy, imbalanced data

Researchers have developed a new framework called CARE to improve machine learning models trained on datasets with both imbalanced class distributions and noisy labels. This method uses insights from vision-language models to adaptively correct errors, applying stricter correction for less frequent classes and more lenient correction for common classes. Experiments show CARE can achieve up to a 3.0% performance improvement over existing techniques. AI

Summary written by gemini-2.5-flash-lite from 2 sources. How we write summaries →

IMPACT Enhances model robustness for real-world datasets, potentially improving performance in applications with skewed data distributions.

RANK_REASON The cluster contains an academic paper detailing a new machine learning framework.

Read on arXiv cs.CV →

COVERAGE [2]

  1. arXiv cs.CV TIER_1 · Mengke Li, Haiquan Ling, Lihao Chen, Yang Lu, Yiqun Zhang, Hui Huang ·

    CARE: Class-Adaptive Expert Consensus for Reliable Learning with Long-Tailed Noisy Labels

    arXiv:2605.23254v1 Announce Type: new Abstract: 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 …

  2. arXiv cs.CV TIER_1 · Hui Huang ·

    CARE: Class-Adaptive Expert Consensus for Reliable Learning with Long-Tailed Noisy Labels

    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 ineffec…