SPLIT allows us to disentangle local and global information into two separate sets of latent variables within the variational autoencoder (VAE) framework . SPLIT adds generative assumption to the VAE by requiring a subset of the latent variables to generate an auxiliary set of observable data . This additional generative assumptions primes the latent variable to local information and encourages the other latent variables to represent global information . The code for SPLIT is at https://://://github.com/51616/split-vae and the code for our experiments is at http://www.cnn.org/suspective-evaluation-of-evaluated/split. SPLIT and recent research in cognitive neuroscience regarding the disentanglement in human visual perception. We establish connections between SPLIT

Links: PDF - Abstract

Code :


Keywords : split - latent - local - global - information -

Leave a Reply

Your email address will not be published. Required fields are marked *