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 Tools annual

Global: Zendy Free Plan

Global: Zendy Tools monthly

Global: Zendy Plus monthly

Affective Computing has recently attracted the attention of the researchcommunity, due to its numerous applications in diverse areas. In this context,the emergence of video-based data allows to enrich the widely used spatialfeatures with the inclusion of temporal information. However, suchspatio-temporal modelling often results in very high-dimensional feature spacesand large volumes of data, making training difficult and time consuming. Thispaper addresses these shortcomings by proposing a novel model that efficientlyextracts both spatial and temporal features of the data by means of itsenhanced temporal modelling based on latent features. Our proposed modelconsists of three major networks, coined Generator, Discriminator, andCombiner, which are trained in an adversarial setting combined with curriculumlearning to enable our adaptive attention modules. In our experiments, we showthe effectiveness of our approach by reporting our competitive results on boththe AFEW-VA and SEWA datasets, suggesting that temporal modelling improves theaffect estimates both in qualitative and quantitative terms. Furthermore, wefind that the inclusion of attention mechanisms leads to the highest accuracyimprovements, as its weights seem to correlate well with the appearance offacial movements, both in terms of temporal localisation and intensity.Finally, we observe the sequence length of around 160\,ms to be the optimum onefor temporal modelling, which is consistent with other relevant findingsutilising similar lengths.

An Enhanced Adversarial Network with Combined Latent Features for Spatio-Temporal Facial Affect Estimation in the Wild