MNIST and SVHN datasets for incomplete image processing

We investigate the problem of training neural networks from incomplete images without replacing missing values.

Data and Resources

Cite this as

Tomasz Danel, Marek Smieja, Lukasz Struski, Przemyslaw Spurek, Lukasz Maziarka (2025). Dataset: MNIST and SVHN datasets for incomplete image processing. https://doi.org/10.57702/iupw9ds8

DOI retrieved: January 3, 2025

Additional Info

Field Value
Created January 3, 2025
Last update January 3, 2025
Defined In https://doi.org/10.48550/arXiv.2010.13914
Author Tomasz Danel
More Authors
Marek Smieja
Lukasz Struski
Przemyslaw Spurek
Lukasz Maziarka