An Acceleration Framework for High Resolution Image Synthesis

TL;DR

This paper introduces a two-stage acceleration framework for high resolution image synthesis with GANs, significantly reducing training time and hardware requirements by operating in a compressed latent space.

Contribution

The proposed method accelerates high resolution image synthesis training by transforming images into a smaller latent space and training a code generator, enabling high-quality results with limited hardware.

Findings

Training time reduced to 3 days for 1024x1024 images

High quality of generated images maintained or improved

Requires only one high-end GPU for large images

Abstract

Synthesis of high resolution images using Generative Adversarial Networks (GANs) is challenging, which usually requires numbers of high-end graphic cards with large memory and long time of training. In this paper, we propose a two-stage framework to accelerate the training process of synthesizing high resolution images. High resolution images are first transformed to small codes via the trained encoder and decoder networks. The code in latent space is times smaller than the original high resolution images. Then, we train a code generation network to learn the distribution of the latent codes. In this way, the generator only learns to generate small latent codes instead of large images. Finally, we decode the generated latent codes to image space via the decoder networks so as to output the synthesized high resolution images. Experimental results show that the proposed method accelerates the training process significantly and increases the quality of the generated samples. The proposed acceleration framework makes it possible to generate high resolution images using less training time with limited hardware resource. After using the proposed acceleration method, it takes only 3 days to train a 1024 *1024 image generator on Celeba-HQ dataset using just one NVIDIA P100 graphic card.