Photographic Image Synthesis with Cascaded Refinement Networks

TL;DR

This paper introduces a non-adversarial, feedforward neural network method for synthesizing high-resolution photographic images from semantic layouts, outperforming existing approaches in realism.

Contribution

It presents a novel end-to-end trained network that generates photorealistic images from semantic maps without adversarial training, scalable to 2-megapixel resolution.

Findings

Synthesizes high-resolution images at 2 megapixels.

Produces more realistic images than alternative methods.

Operates with a single feedforward network trained end-to-end.

Abstract

We present an approach to synthesizing photographic images conditioned on semantic layouts. Given a semantic label map, our approach produces an image with photographic appearance that conforms to the input layout. The approach thus functions as a rendering engine that takes a two-dimensional semantic specification of the scene and produces a corresponding photographic image. Unlike recent and contemporaneous work, our approach does not rely on adversarial training. We show that photographic images can be synthesized from semantic layouts by a single feedforward network with appropriate structure, trained end-to-end with a direct regression objective. The presented approach scales seamlessly to high resolutions; we demonstrate this by synthesizing photographic images at 2-megapixel resolution, the full resolution of our training data. Extensive perceptual experiments on datasets of outdoor and indoor scenes demonstrate that images synthesized by the presented approach are considerably more realistic than alternative approaches. The results are shown in the supplementary video at https://youtu.be/0fhUJT21-bs