Figure 3: Image classification accuracy when the underlying class distribution shifts every 1k iterations. The vertical dashed line indicates the end of an entire cycle through the tasks, and return to the original classification task at t = 0. Left: vanilla classifier. Right: classifier with an additional self-supervised loss. This example demonstrates that a classifier may fail to learn generalizable representations in a non-stationary environment, but self-supervision helps mitigate this problem.
Discriminator with rotation-based self-supervision.
Figure 1: The discriminator D performs two tasks: true vs. fake binary classification, and rotation degree classification. Both the fake and real images are rotated by 0, 90, 180, and 270 degrees. The colored arrows indicate that only the upright(垂直的) images are considered for true vs fake classification loss task. For the rotation loss, all images are classified by the discriminator according to their rotation degree.