Part of Advances in Neural Information Processing Systems 35 (NeurIPS 2022) Main Conference Track
Yossi Gandelsman, Yu Sun, Xinlei Chen, Alexei Efros
Test-time training adapts to a new test distribution on the fly by optimizing a model for each test input using self-supervision.In this paper, we use masked autoencoders for this one-sample learning problem.Empirically, our simple method improves generalization on many visual benchmarks for distribution shifts.Theoretically, we characterize this improvement in terms of the bias-variance trade-off.