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

We propose a discrimination-aware learning method to improve both accuracyand fairness of biased face recognition algorithms. The most popular facerecognition benchmarks assume a distribution of subjects without paying muchattention to their demographic attributes. In this work, we perform acomprehensive discrimination-aware experimentation of deep learning-based facerecognition. We also propose a general formulation of algorithmicdiscrimination with application to face biometrics. The experiments includetree popular face recognition models and three public databases composed of64,000 identities from different demographic groups characterized by gender andethnicity. We experimentally show that learning processes based on the mostused face databases have led to popular pre-trained deep face models thatpresent a strong algorithmic discrimination. We finally propose adiscrimination-aware learning method, Sensitive Loss, based on the populartriplet loss function and a sensitive triplet generator. Our approach works asan add-on to pre-trained networks and is used to improve their performance interms of average accuracy and fairness. The method shows results comparable tostate-of-the-art de-biasing networks and represents a step forward to preventdiscriminatory effects by automatic systems.

SensitiveLoss: Improving Accuracy and Fairness of Face Representations with Discrimination-Aware Deep Learning