论文标题
通过熵正规化在分类中保留细粒特征信息
Preserving Fine-Grain Feature Information in Classification via Entropic Regularization
论文作者
论文摘要
标记分类数据集意味着定义类和相关的粗标签，这可能会近似更光滑，更复杂的地面真相。例如，自然图像可能包含多个对象，其中一个对象在许多视觉数据集中标记，或者可以是由于回归问题的离散化而导致的。在此类粗糙标签上使用跨凝结训练分类模型可能会大致切入特征空间，可能会忽略最有意义的此类功能，特别是在基础细粒任务上丢失了信息。在本文中，我们对仅使用粗粒标签训练的模型解决了解决细粒分类或回归的问题感兴趣。我们表明，标准的跨凝结可能导致与粗相关的特征过度拟合。我们引入了基于熵的正则化，以促进训练有素的模型的特征空间中的更多多样性，并从经验上证明了这种方法的功效，以在细粒度问题上达到更好的性能。通过理论发展和经验验证，我们的结果得到了支持。
Labeling a classification dataset implies to define classes and associated coarse labels, that may approximate a smoother and more complicated ground truth. For example, natural images may contain multiple objects, only one of which is labeled in many vision datasets, or classes may result from the discretization of a regression problem. Using cross-entropy to train classification models on such coarse labels is likely to roughly cut through the feature space, potentially disregarding the most meaningful such features, in particular losing information on the underlying fine-grain task. In this paper we are interested in the problem of solving fine-grain classification or regression, using a model trained on coarse-grain labels only. We show that standard cross-entropy can lead to overfitting to coarse-related features. We introduce an entropy-based regularization to promote more diversity in the feature space of trained models, and empirically demonstrate the efficacy of this methodology to reach better performance on the fine-grain problems. Our results are supported through theoretical developments and empirical validation.
