Intelligent facial emotion recognition and semantic-based topic detection for a humanoid robot
作者：
Highlights：
• The work employs anatomical truth of facial behaviour to recognise seven emotions.
• 17 Action Units are recognised using upper and lower action analysers.
• A facial emotion recogniser uses the derived actions to further decode emotions.
• Latent Semantic Analysis is used to identify topics beyond linguistic constraints.
• The overall development is integrated with a modern humanoid robot platform.
摘要
•The work employs anatomical truth of facial behaviour to recognise seven emotions.•17 Action Units are recognised using upper and lower action analysers.•A facial emotion recogniser uses the derived actions to further decode emotions.•Latent Semantic Analysis is used to identify topics beyond linguistic constraints.•The overall development is integrated with a modern humanoid robot platform.
论文关键词：Facial emotion detection,Action Units,Latent Semantic Analysis,Human robot interaction
论文评审过程：Available online 16 March 2013.
论文官网地址：https://doi.org/10.1016/j.eswa.2013.03.016
