Image inpainting using frequency domain priors

TL;DR

This paper introduces a new image inpainting method that leverages frequency domain priors, specifically the Discrete Fourier Transform, to improve high-frequency detail reconstruction and reduce artifacts.

Contribution

It proposes a frequency-based deconvolution module that enhances neural networks' ability to learn global context and reconstruct high-frequency details in images.

Findings

Outperforms state-of-the-art methods on CelebA, Paris Streetview, and DTD datasets.

Achieves better qualitative and quantitative inpainting results.

Reduces artifacts and improves texture and boundary reconstruction.

Abstract

In this paper, we present a novel image inpainting technique using frequency domain information. Prior works on image inpainting predict the missing pixels by training neural networks using only the spatial domain information. However, these methods still struggle to reconstruct high-frequency details for real complex scenes, leading to a discrepancy in color, boundary artifacts, distorted patterns, and blurry textures. To alleviate these problems, we investigate if it is possible to obtain better performance by training the networks using frequency domain information (Discrete Fourier Transform) along with the spatial domain information. To this end, we propose a frequency-based deconvolution module that enables the network to learn the global context while selectively reconstructing the high-frequency components. We evaluate our proposed method on the publicly available datasets CelebA, Paris Streetview, and DTD texture dataset, and show that our method outperforms current state-of-the-art image inpainting techniques both qualitatively and quantitatively.