Glow: Generative Flow with Invertible 1x1 Convolutions

Published in

Neural Information Processing Systems(2018)

External Links:

Generate Graph

TL;DR

Glow, a simple type of generative flow using an invertible 1x1 convolution, is proposed, demonstrating that a generative model optimized towards the plain log-likelihood objective is capable of efficient realistic-looking synthesis and manipulation of large images.

Abstract

Flow-based generative models (Dinh et al., 2014) are conceptually attractive due to tractability of the exact log-likelihood, tractability of exact latent-variable inference, and parallelizability of both training and synthesis. In this paper we propose Glow, a simple type of generative flow using an invertible 1x1 convolution. Using our method we demonstrate a significant improvement in log-likelihood on standard benchmarks. Perhaps most strikingly, we demonstrate that a generative model optimized towards the plain log-likelihood objective is capable of efficient realistic-looking synthesis and manipulation of large images. The code for our model is available at this https URL

Authors

Diederik P. Kingma

8 papers

Prafulla Dhariwal

8 papers

References29 items

Image Transformer

Parallel WaveNet: Fast High-Fidelity Speech Synthesis

Progressive Growing of GANs for Improved Quality, Stability, and Variation

The Reversible Residual Network: Backpropagation Without Storing Activations

Flow-GAN: Combining Maximum Likelihood and Adversarial Learning in Generative Models

Glow: Generative Flow with Invertible 1x1 Convolutions

Published in

Neural Information Processing Systems(2018)

External Links:

Generate Graph

TL;DR

Abstract

Authors

Diederik P. Kingma

8 papers

Prafulla Dhariwal

8 papers

References29 items

Image Transformer

Parallel WaveNet: Fast High-Fidelity Speech Synthesis

Progressive Growing of GANs for Improved Quality, Stability, and Variation

The Reversible Residual Network: Backpropagation Without Storing Activations

Flow-GAN: Combining Maximum Likelihood and Adversarial Learning in Generative Models

Masked Autoregressive Flow for Density Estimation

Parallel Multiscale Autoregressive Density Estimation

WaveNet: A Generative Model for Raw Audio

Conditional Image Generation with PixelCNN Decoders

Improved Variational Inference with Inverse Autoregressive Flow

Density estimation using Real NVP

Identity Mappings in Deep Residual Networks

Weight Normalization: A Simple Reparameterization to Accelerate Training of Deep Neural Networks

Pixel Recurrent Neural Networks

LSUN: Construction of a Large-scale Image Dataset using Deep Learning with Humans in the Loop

Variational Inference with Normalizing Flows

Batch Normalization: Accelerating Deep Network Training by Reducing Internal Covariate Shift

Adam: A Method for Stochastic Optimization

NICE: Non-linear Independent Components Estimation

ImageNet Large Scale Visual Recognition Challenge

Auto-Encoding Variational Bayes

Generating Sequences With Recurrent Neural Networks

Long Short-Term Memory

GENERATIVE ADVERSARIAL NETS

Variational autoencoders. Under Review

Gradient checkpointing

Learning Multiple Layers of Features from Tiny Images

Higher Order Statistical Decorrelation without Information Loss

Class conditional 32 × 32 ImageNet samples Figure 10: Class conditional samples on 5-bit CIFAR-10 and 32 × 32 ImageNet respectively

Field of Study

Computer ScienceMathematics

Journal Information

Name

ArXiv

Volume

abs/2005.00687

Venue Information

Name

Neural Information Processing Systems

Type

conference

URL

http://neurips.cc/

Alternate Names

Neural Inf Process Syst
NeurIPS
NIPS