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

Complex emotion recognition is a cognitive task that has so far eluded thesame excellent performance of other tasks that are at or above the level ofhuman cognition. Emotion recognition through facial expressions is particularlydifficult due to the complexity of emotions expressed by the human face. For amachine to approach the same level of performance in complex facial expressionrecognition as a human, it may need to synthesise knowledge and understand newconcepts in real-time, as humans do. Humans are able to learn new conceptsusing only few examples by distilling important information from memories.Inspired by human cognition and learning, we propose a novel continual learningmethod for complex facial expression recognition that can accurately recognisenew compound expression classes using few training samples, by building on andretaining its knowledge of basic expression classes. In this work, we also useGradCAM visualisations to demonstrate the relationship between basic andcompound facial expressions. Our method leverages this relationship throughknowledge distillation and a novel Predictive Sorting Memory Replay, to achievethe current state-of-the-art in continual learning for complex facialexpression recognition, with 74.28% Overall Accuracy on new classes. We alsodemonstrate that using continual learning for complex facial expressionrecognition achieves far better performance than non-continual learningmethods, improving on state-of-the-art non-continual learning methods by13.95%. Our work is also the first to apply few-shot learning to complex facialexpression recognition, achieving the state-of-the-art with 100% accuracy usingonly a single training sample per class.

Complex Facial Expression Recognition Using Deep Knowledge Distillation of Basic Features