Normalizing Flows with Multi-Scale Autoregressive Priors

TL;DR

This paper enhances flow-based generative models by integrating multi-scale autoregressive priors to better capture long-range dependencies, resulting in improved density estimation and image generation quality on standard datasets.

Contribution

Introduces multi-scale autoregressive priors for flow models, significantly increasing their expressiveness and performance in image synthesis tasks.

Findings

Achieves state-of-the-art density estimation on MNIST, CIFAR-10, and ImageNet.

Improves image generation quality with higher FID and Inception scores.

Demonstrates better modeling of complex multimodal data.

Abstract

Flow-based generative models are an important class of exact inference models that admit efficient inference and sampling for image synthesis. Owing to the efficiency constraints on the design of the flow layers, e.g. split coupling flow layers in which approximately half the pixels do not undergo further transformations, they have limited expressiveness for modeling long-range data dependencies compared to autoregressive models that rely on conditional pixel-wise generation. In this work, we improve the representational power of flow-based models by introducing channel-wise dependencies in their latent space through multi-scale autoregressive priors (mAR). Our mAR prior for models with split coupling flow layers (mAR-SCF) can better capture dependencies in complex multimodal data. The resulting model achieves state-of-the-art density estimation results on MNIST, CIFAR-10, and ImageNet. Furthermore, we show that mAR-SCF allows for improved image generation quality, with gains in FID and Inception scores compared to state-of-the-art flow-based models.