STL-10

The 'STL-10' dataset is a benchmark that contains 10 classes of images with 5,000 labeled training images and 8,000 unlabeled test images, used to evaluate the performance of machine learning algorithms in unsupervised and semi-supervised learning settings.

Papers using STL-10 (2)

Distilling Knowledge from Self-Supervised Teacher by Embedding Graph Alignment2022 · 2 citesChebLieNet: Invariant Spectral Graph NNs Turned Equivariant by Riemannian Geometry on Lie Groups2021