Identifying through Flows for Recovering Latent Representations

Preprint English OPEN
Li, Shen; Hooi, Bryan; Lee, Gim Hee;
  • Subject: Computer Science - Machine Learning | Statistics - Machine Learning

Identifiability, or recovery of the true latent representations from which the observed data originates, is de facto a fundamental goal of representation learning. Yet, most deep generative models do not address the question of identifiability, and thus fail to deliver ... View more
Share - Bookmark