Learning with noisy labels aims to ensure model generalization given a label-corrupted training set. The sample selection strategy achieves promising performance by selecting a label-reliable subset for model training. In this paper, we empirically reveal that existing sample selection methods suffer from both data and training bias that are represented as imbalanced selected sets and accumulation errors in practice, respectively. However, only the training bias was handled in previous studies. To address this limitation, we propose a noIse-Tolerant Expert Model (ITEM) for debiased learning in sample selection. Specifically, to mitigate the training bias, we design a robust network architecture that integrates with multiple experts. Compared with the prevailing double-branch network, our network exhibits better performance of selection and prediction by ensembling these experts while training with fewer parameters. Meanwhile, to mitigate the data bias, we propose a mixed sampling strategy based on two weight-based data samplers. By training on the mixture of two class-discriminative mini-batches, the model mitigates the effect of the imbalanced training set while avoiding sparse representations that are easily caused by sampling strategies. Extensive experiments and analyses demonstrate the effectiveness of ITEM. Our code is available at this url \href{https://github.com/1998v7/ITEM}{ITEM}.

翻译：针对含噪声标签的学习问题，旨在确保模型在标签受损的训练集上仍能泛化。样本选择策略通过选取标签可靠的子集进行模型训练，取得了显著性能。本文通过实验揭示，现有样本选择方法同时受数据和训练偏差影响——实践中分别表现为不平衡的选定集和累积误差。然而，以往研究仅处理了训练偏差。为突破这一局限，我们提出了一种噪声容忍专家模型（ITEM），用于样本选择中的去偏学习。具体而言，为缓解训练偏差，我们设计了集成多专家的鲁棒网络架构。与主流双分支网络相比，该网络通过集成专家在训练参数更少的同时，展现出更优的选取与预测性能。同时，为缓解数据偏差，我们提出了基于两种权重数据采样器的混合采样策略。通过训练两类判别性小批量的混合数据，模型在避免采样策略易导致的稀疏表征的同时，减轻了不平衡训练集的影响。大量实验与分析验证了ITEM的有效性。我们的代码开源于 \href{https://github.com/1998v7/ITEM}{ITEM}。