论文标题
预测几个射击分类器的准确性
Predicting the Accuracy of a Few-Shot Classifier
论文作者
论文摘要
在几乎没有学习的情况下，由于标记的样本数量少，因此无法使用验证集测量训练有素的分类器的概括能力。在本文中，我们有兴趣寻找替代方案来回答以下问题：我的分类器是否很好地推广到以前看不见的数据？我们首先分析概括性能变异性的原因。然后，我们研究了使用基于转移的解决方案的情况，并考虑三个设置：i）在只能访问一些标记的样本的地方，ii）半监督，我们可以访问一些标记的未标记样本，而III则可以访问我们只能访问未标记的样品。对于每种设置，我们提出了合理的措施，我们从经验上证明，与所考虑的分类器的概括能力相关。我们还表明，这些简单的措施可用于预测泛化，直到某种信心。我们对标准的几片视觉数据集进行了实验。
In the context of few-shot learning, one cannot measure the generalization ability of a trained classifier using validation sets, due to the small number of labeled samples. In this paper, we are interested in finding alternatives to answer the question: is my classifier generalizing well to previously unseen data? We first analyze the reasons for the variability of generalization performances. We then investigate the case of using transfer-based solutions, and consider three settings: i) supervised where we only have access to a few labeled samples, ii) semi-supervised where we have access to both a few labeled samples and a set of unlabeled samples and iii) unsupervised where we only have access to unlabeled samples. For each setting, we propose reasonable measures that we empirically demonstrate to be correlated with the generalization ability of considered classifiers. We also show that these simple measures can be used to predict generalization up to a certain confidence. We conduct our experiments on standard few-shot vision datasets.
