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 Tools annual

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 Tools monthly

Global: Zendy Plus monthly

Presents the access of premium content as premium feature

Premium Content

Global: Zendy Plus annual

Global: Zendy Free Plan

Most modern face completion approaches adopt an autoencoder or its variantsto restore missing regions in face images. Encoders are often utilized to learnpowerful representations that play an important role in meeting the challengesof sophisticated learning tasks. Specifically, various kinds of masks are oftenpresented in face images in the wild, forming complex patterns, especially inthis hard period of COVID-19. It's difficult for encoders to capture suchpowerful representations under this complex situation. To address thischallenge, we propose a self-supervised Siamese inference network to improvethe generalization and robustness of encoders. It can encode contextualsemantics from full-resolution images and obtain more discriminativerepresentations. To deal with geometric variations of face images, a densecorrespondence field is integrated into the network. We further propose amulti-scale decoder with a novel dual attention fusion module (DAF), which cancombine the restored and known regions in an adaptive manner. This multi-scalearchitecture is beneficial for the decoder to utilize discriminativerepresentations learned from encoders into images. Extensive experimentsclearly demonstrate that the proposed approach not only achieves more appealingresults compared with state-of-the-art methods but also improves theperformance of masked face recognition dramatically.

Contrastive Attention Network with Dense Field Estimation for Face Completion