Multiview facial expression recognition with Mapping approaches

Artificial Intelligence (AI) and Human Computer Interaction (HCI) have increasingly attracted attention. There are various tasks regarding to the human and human faces handled with computer vision in terms of recognition. A new challenge and an extended form of facial expression recognition (FER) is Multiview Facial Expression Recognition (MFER) which is able to recognize human emotions even from non-frontal faces. Indeed, it is an unrealistic assumption if we assume that the human faces are always frontal (especially in the wild), therefore, developing new ideas that can improve FER systems to MFER is always desirable. However, this is not a naive task due to the lack of human face parts in non-frontal view. In this book we address this problem by compensating unavailable parts using mapping techniques and an inpainting approach.

Autorentext

Mahdi Jampour is assistant professor at Quchan University of Technology. He received his PhD in Computer Vision from Graz University of Technology. He achieved the Best MSc thesis award, full PhD scholarship, Postdoctoral fellow and also awarded Marshal Plan Fellowship to visit the University of Massachusetts. He works in area of computer vision.