Abstract
In the image classification task, deep neural networks frequently rely on bias attributes that are spuriously correlated with a target class in the presence of dataset bias, resulting in degraded performance when applied to data without bias attributes. The task of debiasing aims to compel classifiers to learn intrinsic attributes that inherently define a target class rather than focusing on bias attributes. While recent approaches mainly focus on emphasizing the learning of data samples without bias attributes (i.e., bias-conflicting samples) compared to samples with bias attributes (i.e., bias-aligned samples), they fall short of directly guiding models where to focus for learning intrinsic features. To address this limitation, this paper proposes a method that provides the model with explicit spatial guidance that indicates the region of intrinsic features. We first identify the intrinsic features by investigating the class-discerning common features between a bias-aligned (BA) sample and a bias-conflicting (BC) sample (i.e., bias-contrastive pair). Next, we enhance the intrinsic features in the BA sample that are relatively under-exploited for prediction compared to the BC sample. To construct the bias-contrastive pair without using bias information, we introduce a bias-negative score that distinguishes BC samples from BA samples employing a biased model. The experiments demonstrate that our method achieves state-of-the-art performance on synthetic and real-world datasets with various levels of bias severity.
Abstract (translated)
在图像分类任务中，深度神经网络经常依赖具有 dataset bias 中的目标类偏见属性的偏见属性，从而在应用无偏见数据时导致性能下降。去偏任务旨在迫使分类器学习固有属性，而不是将重点放在偏见属性上。尽管最近的方法主要关注强调无偏见属性数据样本的学习（即 bias-conflicting 样本）与有偏见属性样本（即 bias-aligned 样本）之间的差异，但它们未能直接指导模型在何处集中学习固有特征。为了克服这一局限，本文提出了一种为模型提供明确的空间指导的方法，该指导表示固有特征的区域。我们首先通过研究偏重（BA）样本和有偏见（BC）样本之间的类区分共同特征来确定固有特征。接下来，我们在BA样本中增强那些相对于预测相对较少被利用的固有特征。为了在没有偏见信息的情况下构建偏置对比对，我们引入了一个基于有偏模型的偏见负分数，用于将BC样本与BA样本区分开来。实验证明，我们的方法在具有各种程度偏差严重性的合成和真实世界数据上实现了最先进的性能。
URL
https://arxiv.org/abs/2404.19250
