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

Robots and artificial agents that interact with humans should be able to doso without bias and inequity, but facial perception systems have notoriouslybeen found to work more poorly for certain groups of people than others. In ourwork, we aim to build a system that can perceive humans in a more transparentand inclusive manner. Specifically, we focus on dynamic expressions on thehuman face, which are difficult to collect for a broad set of people due toprivacy concerns and the fact that faces are inherently identifiable.Furthermore, datasets collected from the Internet are not necessarilyrepresentative of the general population. We address this problem by offering aSim2Real approach in which we use a suite of 3D simulated human models thatenables us to create an auditable synthetic dataset covering 1)underrepresented facial expressions, outside of the six basic emotions, such asconfusion; 2) ethnic or gender minority groups; and 3) a wide range of viewingangles that a robot may encounter a human in the real world. By augmenting asmall dynamic emotional expression dataset containing 123 samples with asynthetic dataset containing 4536 samples, we achieved an improvement inaccuracy of 15% on our own dataset and 11% on an external benchmark dataset,compared to the performance of the same model architecture without synthetictraining data. We also show that this additional step improves accuracyspecifically for racial minorities when the architecture's feature extractionweights are trained from scratch.

Towards Inclusive HRI: Using Sim2Real to Address Underrepresentation in Emotion Expression Recognition