<ul><li>The study focuses on understanding and mitigating bias in sample selection for learning with noisy labels.</li><li>Existing sample selection methods suffer from both data and training bias.</li><li>The research introduces the NoIse-Tolerant Expert Model (ITEM) to address the limitations.</li><li>ITEM incorporates a robust network architecture and a mixed sampling strategy to mitigate both training and data bias.</li></ul>

Understanding and Mitigating the Bias in Sample Selection for Learning with Noisy Labels

Discover more