Scene Graph to Image Generation with Contextualized Object Layout Refinement

TL;DR

This paper introduces a novel method for scene graph to image generation that progressively refines object layouts to enhance coverage and reduce overlap, resulting in higher quality images.

Contribution

It proposes a new approach that generates object layouts gradually, improving inter-object dependency and layout quality over prior independent methods.

Findings

Layout coverage improved by nearly 20 points

Object overlap reduced to negligible levels

Enhanced image quality demonstrated on COCO-STUFF dataset

Abstract

Generating images from scene graphs is a challenging task that attracted substantial interest recently. Prior works have approached this task by generating an intermediate layout description of the target image. However, the representation of each object in the layout was generated independently, which resulted in high overlap, low coverage, and an overall blurry layout. We propose a novel method that alleviates these issues by generating the entire layout description gradually to improve inter-object dependency. We empirically show on the COCO-STUFF dataset that our approach improves the quality of both the intermediate layout and the final image. Our approach improves the layout coverage by almost 20 points and drops object overlap to negligible amounts.