English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

The purpose of image inpainting is to recover scratches and damaged areasusing context information from remaining parts. In recent years, thanks to theresurgence of convolutional neural networks (CNNs), image inpainting task hasmade great breakthroughs. However, most of the work consider insufficient typesof mask, and their performance will drop dramatically when encountering unseenmasks. To combat these challenges, we propose a simple yet general method tosolve this problem based on the LaMa image inpainting framework, dubbed GLaMa.Our proposed GLaMa can better capture different types of missing information byusing more types of masks. By incorporating more degraded images in thetraining phase, we can expect to enhance the robustness of the model withrespect to various masks. In order to yield more reasonable results, we furtherintroduce a frequency-based loss in addition to the traditional spatialreconstruction loss and adversarial loss. In particular, we introduce aneffective reconstruction loss both in the spatial and frequency domain toreduce the chessboard effect and ripples in the reconstructed image. Extensiveexperiments demonstrate that our method can boost the performance over theoriginal LaMa method for each type of mask on FFHQ, ImageNet, Places2 andWikiArt dataset. The proposed GLaMa was ranked first in terms of PSNR, LPIPSand SSIM in the NTIRE 2022 Image Inpainting Challenge Track 1 Unsupervised.

GLaMa: Joint Spatial and Frequency Loss for General Image Inpainting