George Washington University
Unsupervised Prejudice: Image classification models learned bias from ImageNet.
Social biases are well documented in decisions made by supervised models trained on ImageNet’s labels. But they also crept into the output of unsupervised models pretrained on the same dataset.