A comparative study of attention mechanism and generative adversarial network in facade damage segmentation

TL;DR

This paper compares the effectiveness of attention mechanisms and generative adversarial networks within U-net architectures for facade damage segmentation, revealing their individual and combined impacts on segmentation quality.

Contribution

It provides a systematic comparison of attention and GAN strategies in U-net for facade damage segmentation, including their combined effects.

Findings

Attention mechanism improves segmentation accuracy.

GAN enhances the realism of segmented images.

Combined approach yields the best performance.

Abstract

Semantic segmentation profits from deep learning and has shown its possibilities in handling the graphical data from the on-site inspection. As a result, visual damage in the facade images should be detected. Attention mechanism and generative adversarial networks are two of the most popular strategies to improve the quality of semantic segmentation. With specific focuses on these two strategies, this paper adopts U-net, a representative convolutional neural network, as the primary network and presents a comparative study in two steps. First, cell images are utilized to respectively determine the most effective networks among the U-nets with attention mechanism or generative adversarial networks. Subsequently, selected networks from the first test and their combination are applied for facade damage segmentation to investigate the performances of these networks. Besides, the combined effect of the attention mechanism and the generative adversarial network is discovered and discussed.