Vehicle Image Generation Going Well with The Surroundings

TL;DR

This paper introduces a novel neural network approach for generating realistic vehicle images in real-world scenes without needing segmentation layouts, improving quality through shape and detail refinement, and validated on popular datasets.

Contribution

A new vehicle image generation method that does not require segmentation layouts, utilizing shape and detail refinement subnetworks for higher quality images.

Findings

Effective vehicle image generation in real-world scenes.

High-quality images validated by object detection and FID scores.

Applicable to datasets like BDD and Cityscape.

Abstract

Since the generative neural networks have made a breakthrough in the image generation problem, lots of researches on their applications have been studied such as image restoration, style transfer and image completion. However, there has been few research generating objects in uncontrolled real-world environments. In this paper, we propose a novel approach for vehicle image generation in real-world scenes. Using a subnetwork based on a precedent work of image completion, our model makes the shape of an object. Details of objects are trained by an additional colorization and refinement subnetwork, resulting in a better quality of generated objects. Unlike many other works, our method does not require any segmentation layout but still makes a plausible vehicle in the image. We evaluate our method by using images from Berkeley Deep Drive (BDD) and Cityscape datasets, which are widely used for object detection and image segmentation problems. The adequacy of the generated images by the proposed method has also been evaluated using a widely utilized object detection algorithm and the FID score.