Saved in:
Bibliographic Details
Main Authors: Kiani, Kourosh, Rastgoo, Razieh, Chaji, Alireza, Escalera, Sergio
Format: Preprint
Published: 2024
Subjects:
Online Access:https://arxiv.org/abs/2411.05705
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866912111632842752
author Kiani, Kourosh
Rastgoo, Razieh
Chaji, Alireza
Escalera, Sergio
author_facet Kiani, Kourosh
Rastgoo, Razieh
Chaji, Alireza
Escalera, Sergio
contents Image inpainting, the process of restoring missing or corrupted regions of an image by reconstructing pixel information, has recently seen considerable advancements through deep learning-based approaches. In this paper, we introduce a novel deep learning-based pre-processing methodology for image inpainting utilizing the Vision Transformer (ViT). Our approach involves replacing masked pixel values with those generated by the ViT, leveraging diverse visual patches within the attention matrix to capture discriminative spatial features. To the best of our knowledge, this is the first instance of such a pre-processing model being proposed for image inpainting tasks. Furthermore, we show that our methodology can be effectively applied using the pre-trained ViT model with pre-defined patch size. To evaluate the generalization capability of the proposed methodology, we provide experimental results comparing our approach with four standard models across four public datasets, demonstrating the efficacy of our pre-processing technique in enhancing inpainting performance.
format Preprint
id arxiv_https___arxiv_org_abs_2411_05705
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Image inpainting enhancement by replacing the original mask with a self-attended region from the input image
Kiani, Kourosh
Rastgoo, Razieh
Chaji, Alireza
Escalera, Sergio
Computer Vision and Pattern Recognition
Image and Video Processing
Image inpainting, the process of restoring missing or corrupted regions of an image by reconstructing pixel information, has recently seen considerable advancements through deep learning-based approaches. In this paper, we introduce a novel deep learning-based pre-processing methodology for image inpainting utilizing the Vision Transformer (ViT). Our approach involves replacing masked pixel values with those generated by the ViT, leveraging diverse visual patches within the attention matrix to capture discriminative spatial features. To the best of our knowledge, this is the first instance of such a pre-processing model being proposed for image inpainting tasks. Furthermore, we show that our methodology can be effectively applied using the pre-trained ViT model with pre-defined patch size. To evaluate the generalization capability of the proposed methodology, we provide experimental results comparing our approach with four standard models across four public datasets, demonstrating the efficacy of our pre-processing technique in enhancing inpainting performance.
title Image inpainting enhancement by replacing the original mask with a self-attended region from the input image
topic Computer Vision and Pattern Recognition
Image and Video Processing
url https://arxiv.org/abs/2411.05705