MNIST-M

Introduced by Ganin et al. in Domain-Adversarial Training of Neural Networks

MNIST-M is created by combining MNIST digits with the patches randomly extracted from color photos of BSDS500 as their background. It contains 59,001 training and 90,001 test images.

Source: A Review of Single-Source Deep Unsupervised Visual Domain Adaptation

Papers


Paper Code Results Date Stars

Dataset Loaders


Tasks


Similar Datasets


License


  • Unknown

Modalities


Languages