Test-Time Training with Self-Supervision for Generalization under Distribution Shifts

Yu Sun · Xiaolong Wang · Zhuang Liu · John Miller · Alexei Efros · Moritz Hardt


Keywords: [ Algorithms ] [ Trustworthy Machine Learning ] [ Robust Statistics and Machine Learning ]


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.

Chat is not available.