Section outline
-
Aviv Gabbay (HUJI)Sunday 29/3/20
Title: Demystifying Inter-Class Disentanglement
Abstract:Learning to disentangle the hidden factors of variations within a set of observations is a key task for artificial intelligence. In this talk, I will present the problem of class and content disentanglement and illustrate the limitations of current methods. I will therefore introduce LORD, a novel method based on Latent Optimization for Representation Disentanglement. Latent optimization, along with an asymmetric noise regularization, is shown to be superior to amortized inference for achieving disentangled representations. In extensive experiments, our method is shown to achieve better disentanglement performance than both adversarial and non-adversarial methods that use the same level of supervision. I will further introduce a clustering-based approach for extending our method for settings that exhibit in-class variation with promising results on the task of domain translation.