Deep learning for image inpainting: A survey

Image inpainting has been widely exploited in the field of computer vision and image processing. The main purpose of image inpainting is to produce visually plausible structure and texture for the missing regions of damaged images. In the past decade, the success of deep learning has brought new opportu-nities to many vision tasks, which promoted the development of a large number of deep learning-based image inpainting methods. Although these methods have many similarities, they also have their own characteristics due to the differences in data types, application scenarios, computing platforms, etc. It is necessary to classify and summarize these methods to provide a reference for the research community. In this survey, we present a comprehensive overview of recent advances in deep learning-based image inpainting. First, we categorize the deep learning-based techniques from multiple perspectives: inpaint-ing strategies, network structures, and loss functions. Second, we summarize the open source codes and representative public datasets, and introduce the evaluation metrics for quantitative comparisons. Third, we summarize the real-world applications of image inpainting in different scenarios, and give a detailed analysis on the performance of different inpainting algorithms. At last, we conclude the survey and dis-cuss about the future directions.(c) 2022 Elsevier Ltd. All rights reserved.

Automated summary

This article provides a comprehensive overview of recent advances in deep learning-based image inpainting. It categorizes the techniques based on inpainting strategies, network structures, and loss functions. It also summarizes open source codes, public datasets, and evaluation metrics for quantitative comparisons. Additionally, it discusses the real-world applications of image inpainting and analyzes the performance of different inpainting algorithms. Finally, it concludes the survey and explores future directions.