深度主动学习在图像分类中的有效性实证研究
An Empirical Study on the Efficacy of Deep Active Learning for Image Classification
摘要：
深度主动学习（DAL）被认为是降低监督学习中标记成本的一种有前途的方法。然而，现有的DAL方法的评估是基于不同的设置，其结果是有争议的。为了解决这个问题，本文全面评估了19个现有的DAL方法在一个统一的设置，包括传统的全\underline{s}的\underline{a}主动\underline{l}收益（SAL）的战略和新兴的\underline{s} emi- \underline{s}的\underline{a}主动\underline{l}收益（SSAL）技术。我们有几个重要的发现。首先，大多数SAL方法不能达到比随机选择更高的精度。其次，与纯SAL方法相比，半监督训练带来了显着的性能改善。第三，在SSAL设置中执行数据选择可以实现显著且一致的性能改进，特别是对于大量未标记的数据。我们的研究结果为从业者提供了以下指导：应该（i）尽早应用SSAL，（ii）尽可能收集更多未标记的数据，以获得更好的模型性能。
英文摘要：
Deep Active Learning (DAL) has been advocated as a promising method to reduce labeling costs in supervised learning. However, existing evaluations of DAL methods are based on different settings, and their results are controversial. To tackle this issue, this paper comprehensively evaluates 19 existing DAL methods in a uniform setting, including traditional fully-\underline{s}upervised \underline{a}ctive \underline{l}earning (SAL) strategies and emerging \underline{s}emi-\underline{s}upervised \underline{a}ctive \underline{l}earning (SSAL) techniques. We have several non-trivial findings. First, most SAL methods cannot achieve higher accuracy than random selection. Second, semi-supervised training brings significant performance improvement compared to pure SAL methods. Third, performing data selection in the SSAL setting can achieve a significant and consistent performance improvement, especially with abundant unlabeled data. Our findings produce the following guidance for practitioners: one should (i) apply SSAL early and (ii) collect more unlabeled data whenever possible, for better model performance.
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
