GLocal: Global Graph Reasoning and Local Structure Transfer for Person Image Generation

TL;DR

GLocal introduces a framework for person image generation that globally reasons about style correlations and transfers local structures to improve occlusion handling and pose variation in generated images.

Contribution

It proposes a novel GLocal framework that combines global style reasoning with local structure transfer for enhanced person image generation.

Findings

Improves occlusion-aware texture estimation

Effectively recovers corrupted textures in inpainting

Demonstrates superior performance on DeepFashion dataset

Abstract

In this paper, we focus on person image generation, namely, generating person image under various conditions, e.g., corrupted texture or different pose. To address texture occlusion and large pose misalignment in this task, previous works just use the corresponding region's style to infer the occluded area and rely on point-wise alignment to reorganize the context texture information, lacking the ability to globally correlate the region-wise style codes and preserve the local structure of the source. To tackle these problems, we present a GLocal framework to improve the occlusion-aware texture estimation by globally reasoning the style inter-correlations among different semantic regions, which can also be employed to recover the corrupted images in texture inpainting. For local structural information preservation, we further extract the local structure of the source image and regain it in the generated image via local structure transfer. We benchmark our method to fully characterize its performance on DeepFashion dataset and present extensive ablation studies that highlight the novelty of our method.