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

Surveillance scenarios are prone to several problems since they usuallyinvolve low-resolution footage, and there is no control of how far the subjectsmay be from the camera in the first place. This situation is suitable for theapplication of upsampling (super-resolution) algorithms since they may be ableto recover the discriminant properties of the subjects involved. While generalsuper-resolution approaches were proposed to enhance image quality forhuman-level perception, biometrics super-resolution methods seek the best"computer perception" version of the image since their focus is on improvingautomatic recognition performance. Convolutional neural networks and deeplearning algorithms, in general, have been applied to computer vision tasks andare now state-of-the-art for several sub-domains, including imageclassification, restoration, and super-resolution. However, no work hasevaluated the effects that the latest proposed super-resolution methods mayhave upon the accuracy and face verification performance in low-resolution"in-the-wild" data. This project aimed at evaluating and adapting differentdeep neural network architectures for the task of face super-resolution drivenby face recognition performance in real-world low-resolution images. Theexperimental results in a real-world surveillance and attendance datasetsshowed that general super-resolution architectures might enhance faceverification performance of deep neural networks trained on high-resolutionfaces. Also, since neural networks are function approximators and can betrained based on specific objective functions, the use of a customized lossfunction optimized for feature extraction showed promising results forrecovering discriminant features in low-resolution face images.

Analysis and evaluation of Deep Learning based Super-Resolution algorithms to improve performance in Low-Resolution Face Recognition