Dataset biases are notoriously detrimental to model robustness and generalization. The identify-emphasize paradigm appears to be effective in dealing with unknown biases. However, we discover that it is still plagued by two challenges: A, the quality of the identified bias-conflicting samples is far from satisfactory; B, the emphasizing strategies only produce suboptimal performance. In this paper, for challenge A, we propose an effective bias-conflicting scoring method (ECS) to boost the identification accuracy, along with two practical strategies -- peer-picking and epoch-ensemble. For challenge B, we point out that the gradient contribution statistics can be a reliable indicator to inspect whether the optimization is dominated by bias-aligned samples. Then, we propose gradient alignment (GA), which employs gradient statistics to balance the contributions of the mined bias-aligned and bias-conflicting samples dynamically throughout the learning process, forcing models to leverage intrinsic features to make fair decisions. Furthermore, we incorporate self-supervised (SS) pretext tasks into training, which enable models to exploit richer features rather than the simple shortcuts, resulting in more robust models. Experiments are conducted on multiple datasets in various settings, demonstrating that the proposed solution can mitigate the impact of unknown biases and achieve state-of-the-art performance.

翻译：数据集偏差对模型鲁棒性和泛化能力具有公认的负面影响。识别-强调范式在处理未知偏差方面表现出有效性。然而，我们发现该范式仍面临两大挑战：A，已识别的偏差冲突样本质量远未达到令人满意的水平；B，强调策略仅能产生次优性能。针对挑战A，本文提出一种高效的偏差冲突评分方法（ECS）以提升识别准确率，并辅以两项实用策略——同伴挑选（peer-picking）与周期集成（epoch-ensemble）。针对挑战B，我们指出梯度贡献统计量可作为可靠指标，用于检测优化过程是否由偏差对齐样本主导。进而，我们提出梯度对齐（GA）方法，利用梯度统计在整个学习过程中动态平衡已挖掘的偏差对齐样本与偏差冲突样本的贡献，迫使模型利用内在特征做出公平决策。此外，我们将自监督（SS）预训练任务纳入训练流程，使模型能够利用更丰富的特征而非简单捷径，从而获得更鲁棒的模型。在多种设置下的多数据集实验表明，所提方案可有效减轻未知偏差的影响，并实现了最先进的性能。