Semantically Accurate Super-Resolution Generative Adversarial Networks

Abstract

This work addresses the problems of semantic segmentation and image super-resolution by jointly considering the performance of both in training a Generative Adversarial Network (GAN). We propose a novel architecture and domain-specific feature loss, allowing super-resolution to operate as a pre-processing step to increase the performance of downstream computer vision tasks, specifically semantic segmentation. We demonstrate this approach using Nearmap's aerial imagery dataset which covers hundreds of urban areas at 5-7 cm per pixel resolution. We show the proposed approach improves perceived image quality as well as quantitative segmentation accuracy across all prediction classes, yielding an average accuracy improvement of 11.8% and 108% at 4× and 32× super-resolution, compared with state-of-the art single-network methods. This work demonstrates that jointly considering image-based and task-specific losses can improve the performance of both, and advances the state-of-the-art in semantic-aware super-resolution of aerial imagery.

Super-resolution with generative adversarial networks for improved object detection in aerial images

“…Because the input image does not have a high-resolution pair there is a comparison of the results with some other methods but there is not any performance metric result for the super-resolution on the DOTA data set. Frizza et al (2022) proposed a GAN-based approach for semantic segmentation on aerial images. The work addresses the problems of semantic segmentation and image superresolution by jointly considering the performance of both in training a GAN.…”

“…Frizza et al (2022) proposed a GAN-based approach for semantic segmentation on aerial images. The work addresses the problems of semantic segmentation and image super-resolution by jointly considering the performance of both in training a GAN.…”