This is a demo for the ICLR 2022 submission 'Understanding Latent Correlation-Based
Multiview Learning and Self-Supervision: An Identifiability Perspective'.

Please run the following command to see the reults on the MNIST dataset.
Tested on: python 3.5, PyTorch 1.1.0, cuda 9.0, scikit-learn and Scipy.

bash run_mnist.sh
