Test-Time Training with Self-Supervision for Generalization under Distribution Shifts
Test-Time Training with Self-Supervision for Generalization under Distribution Shifts
复制标题
DOI：
--
复制
发表时间：
2019-09
期刊：
影响因子：
--
通讯作者：
Yu Sun;X. Wang;Zhuang Liu;John Miller;Alexei A. Efros;Moritz Hardt
中科院分区：
文献类型：
--
作者：
Yu Sun;X. Wang;Zhuang Liu;John Miller;Alexei A. Efros;Moritz Hardt
在本文中，我们提出了测试时训练，这是一种当训练和测试数据来自不同分布时提高预测模型性能的通用方法。我们将单个未标记的测试样本转变为自监督学习问题，在进行预测之前更新模型参数。这也自然地扩展到在线流中的数据。我们的简单方法改进了各种图像分类基准，旨在评估分布变化的稳健性。
In this paper, we propose Test-Time Training, a general approach for improving the performance of predictive models when training and test data come from different distributions. We turn a single unlabeled test sample into a self-supervised learning problem, on which we update the model parameters before making a prediction. This also extends naturally to data in an online stream. Our simple approach leads to improvements on diverse image classification benchmarks aimed at evaluating robustness to distribution shifts.