TzK Flow - Conditional Generative Model

TL;DR

TzK Flow is a novel conditional generative model that combines the advantages of flow-based models, VAEs, and autoencoders to efficiently learn data distributions with stable training and flexible conditioning.

Contribution

It introduces TzK, a flow-based conditional model that leverages side information and meta-data, avoiding variational approximations and supporting various learning paradigms.

Findings

Achieves state-of-the-art results on MNIST and Omniglot datasets.

Supports supervised, unsupervised, and semi-supervised learning.

Demonstrates stable training and efficient approximation of data distributions.

Abstract

We introduce TzK (pronounced "task"), a conditional probability flow-based model that exploits attributes (e.g., style, class membership, or other side information) in order to learn tight conditional prior around manifolds of the target observations. The model is trained via approximated ML, and offers efficient approximation of arbitrary data sample distributions (similar to GAN and flow-based ML), and stable training (similar to VAE and ML), while avoiding variational approximations. TzK exploits meta-data to facilitate a bottleneck, similar to autoencoders, thereby producing a low-dimensional representation. Unlike autoencoders, the bottleneck does not limit model expressiveness, similar to flow-based ML. Supervised, unsupervised, and semi-supervised learning are supported by replacing missing observations with samples from learned priors. We demonstrate TzK by training jointly on MNIST and Omniglot datasets with minimal preprocessing, and weak supervision, with results comparable to state-of-the-art.