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

Facial expression recognition (FER) has received increasing interest incomputer vision. We propose the TransFER model which can learn richrelation-aware local representations. It mainly consists of three components:Multi-Attention Dropping (MAD), ViT-FER, and Multi-head Self-Attention Dropping(MSAD). First, local patches play an important role in distinguishing variousexpressions, however, few existing works can locate discriminative and diverselocal patches. This can cause serious problems when some patches are invisibledue to pose variations or viewpoint changes. To address this issue, the MAD isproposed to randomly drop an attention map. Consequently, models are pushed toexplore diverse local patches adaptively. Second, to build rich relationsbetween different local patches, the Vision Transformers (ViT) are used in FER,called ViT-FER. Since the global scope is used to reinforce each local patch, abetter representation is obtained to boost the FER performance. Thirdly, themulti-head self-attention allows ViT to jointly attend to features fromdifferent information subspaces at different positions. Given no explicitguidance, however, multiple self-attentions may extract similar relations. Toaddress this, the MSAD is proposed to randomly drop one self-attention module.As a result, models are forced to learn rich relations among diverse localpatches. Our proposed TransFER model outperforms the state-of-the-art methodson several FER benchmarks, showing its effectiveness and usefulness.

TransFER: Learning Relation-aware Facial Expression Representations with Transformers