论文标题
弱监督的多标签分类中的大损失事项
Large Loss Matters in Weakly Supervised Multi-Label Classification
论文作者
论文摘要
弱监督的多标签分类（WSML）任务是使用每个图像的部分观察到的标签学习多标签分类，由于其巨大的注释成本，它变得越来越重要。在这项工作中，我们首先将未观察到的标签视为负标签，将WSML任务投入到嘈杂的多标签分类中。从这个角度来看，我们从经验上观察到，在嘈杂的多级环境中最初发现的记忆效应也发生在多标签环境中。也就是说，该模型首先了解清洁标签的表示，然后开始记住嘈杂的标签。基于这一发现，我们提出了WSML的新方法，该方法拒绝或纠正大型损失样本，以防止模型记住嘈杂的标签。如果没有沉重且复杂的组件，我们提出的方法在几种部分标签设置上的先前最先前的WSML方法（包括Pascal VOC 2012，Coco，MS Coco，Nuswide，Cub，Cub和OpenImimages V3数据集）都优于先前最先前的WSML方法。各种分析还表明，我们的方法实际上效果很好，验证了在弱监督的多标签分类中正确处理大损失的问题。我们的代码可在https://github.com/snucml/largelossmatters上找到。
Weakly supervised multi-label classification (WSML) task, which is to learn a multi-label classification using partially observed labels per image, is becoming increasingly important due to its huge annotation cost. In this work, we first regard unobserved labels as negative labels, casting the WSML task into noisy multi-label classification. From this point of view, we empirically observe that memorization effect, which was first discovered in a noisy multi-class setting, also occurs in a multi-label setting. That is, the model first learns the representation of clean labels, and then starts memorizing noisy labels. Based on this finding, we propose novel methods for WSML which reject or correct the large loss samples to prevent model from memorizing the noisy label. Without heavy and complex components, our proposed methods outperform previous state-of-the-art WSML methods on several partial label settings including Pascal VOC 2012, MS COCO, NUSWIDE, CUB, and OpenImages V3 datasets. Various analysis also show that our methodology actually works well, validating that treating large loss properly matters in a weakly supervised multi-label classification. Our code is available at https://github.com/snucml/LargeLossMatters.
