Legacy Photo Editing with Learned Noise Prior
This work provides a comprehensive solution for restoring old, degraded photographs, benefiting archivists, historians, and individuals with personal collections by improving the visual quality of historical images.
This paper addresses the challenge of restoring legacy photographs, which are often noisy, incomplete, and grayscale. The authors propose a noise prior learner (NEGAN) to simulate real legacy photo noise distributions and an IEGAN framework for joint denoising, inpainting, and colorization, achieving the best perceptual quality compared to state-of-the-art methods.
There are quite a number of photographs captured under undesirable conditions in the last century. Thus, they are often noisy, regionally incomplete, and grayscale formatted. Conventional approaches mainly focus on one point so that those restoration results are not perceptually sharp or clean enough. To solve these problems, we propose a noise prior learner NEGAN to simulate the noise distribution of real legacy photos using unpaired images. It mainly focuses on matching high-frequency parts of noisy images through discrete wavelet transform (DWT) since they include most of noise statistics. We also create a large legacy photo dataset for learning noise prior. Using learned noise prior, we can easily build valid training pairs by degrading clean images. Then, we propose an IEGAN framework performing image editing including joint denoising, inpainting and colorization based on the estimated noise prior. We evaluate the proposed system and compare it with state-of-the-art image enhancement methods. The experimental results demonstrate that it achieves the best perceptual quality. https://github.com/zhaoyuzhi/Legacy-Photo-Editing-with-Learned-Noise-Prior for the codes and the proposed LP dataset.