Fuzzy emotion: a natural approach to automatic facial expression recognition from psychological perspective using fuzzy system

Research output: Contribution to journalArticle

1 Citation (Scopus)

Abstract

Many studies in automatic facial expression recognitions merely limit their focus on recognizing basic emotions, ignoring the fact that humans show various emotions in their daily life. Moreover, from psychological perspective humans express multiple emotions simultaneously. Up to now, researchers recognize two basic emotions at the same time, called mixed emotions. Nevertheless, the mixed emotion still does not reflect how humans express the emotion naturally. This paper advances the concept of mixed emotion into a generalized fuzzy emotion. Fuzzy emotion captures multiple emotions in a single image using fuzzy inference engine. We propose a fuzzy emotion framework which consists of processing system and knowledge system. The processing system extracts facial expression parameters, and the knowledge system employs a fuzzy knowledge-based engine, elicited from the psychologist knowledge to recognize facial expressions. Some advantages are offered: (1) no facial template comparison; (2) no training efforts needed; (3) moreover, fuzzy emotion can recognize ambiguous facial expressions adaptively. The experiment gives a recognition result with the highest accuracy rate of 0.90. A research agenda for future study of mixed emotion recognition is proposed.

Original languageEnglish
JournalCognitive Processing
DOIs
Publication statusPublished - 1 Jan 2019

Keywords

  • Affective computing
  • Artificial intelligence
  • Emotion recognition
  • Facial expression
  • Fuzzy emotion
  • Fuzzy system

Fingerprint Dive into the research topics of 'Fuzzy emotion: a natural approach to automatic facial expression recognition from psychological perspective using fuzzy system'. Together they form a unique fingerprint.

  • Cite this