Fetching the paper…

Multi-Level Variational Autoencoder: Learning Disentangled Representations from Grouped Observations · Around