Current browse context:
cs.LG
Change to browse by:
References & Citations
Computer Science > Machine Learning
Title: InfoCatVAE: Representation Learning with Categorical Variational Autoencoders
(Submitted on 20 Jun 2018 (v1), last revised 25 Jun 2018 (this version, v2))
Abstract: This paper describes InfoCatVAE, an extension of the variational autoencoder that enables unsupervised disentangled representation learning. InfoCatVAE uses multimodal distributions for the prior and the inference network and then maximizes the evidence lower bound objective (ELBO). We connect the new ELBO derived for our model with a natural soft clustering objective which explains the robustness of our approach. We then adapt the InfoGANs method to our setting in order to maximize the mutual information between the categorical code and the generated inputs and obtain an improved model.
Submission history
From: Edouard Pineau [view email][v1] Wed, 20 Jun 2018 16:56:08 GMT (360kb,D)
[v2] Mon, 25 Jun 2018 14:58:24 GMT (359kb,D)
Link back to: arXiv, form interface, contact.