Abstract

Expression recognition in the wild is easily distorted by nonfrontal and asymmetry faces. In nonfrontal faces, some areas are compressed and distorted. Even after frontalization, these compressed areas may still be blurred and distort expression recognition. Additionally, asymmetrical expressions are common on half or local face areas and produce incorrect expression features. Therefore, this paper presents a half-face frontalization and pyramid Fourier frequency conversion method. Despite the location, range and intensity of incorrect expressions in nonfrontal faces being unknown, according to discrete Fourier transform, it can be proven that the frequency band of the correct expression is much larger than that of incorrect expression on the same face. This can be taken advantage of by pyramid frequency conversion, which is designed based on Fourier frequency conversion. It can adjust incorrect expression frequency in multiscales to take them out off the band-pass of the convolution operations of deep learning and be eliminated completely, whereas correct expression information is reserved. Thus, expressions can be recognized effectively.

Highlights

  • Facial expression recognition (FER) is important for artificial intelligence (AI) [1]

  • When AI assists in our work, FER plays an important role for human state feedback, e.g., teaching effect analysis in network training

  • Nonfrontal face frontalization is the best method for solving this problem; this paper introduces the position map regression network (PRNet) network, which is a deep learning method and can accomplish

Read more

Summary

Introduction

Facial expression recognition (FER) is important for artificial intelligence (AI) [1]. The ideal method should recognize expression in the wild, which means the recognition method should adapt to multiview human faces [2], nonfrontal faces [3]; this has not been achieved [4]. Nonfrontal faces can seriously distort face images [5], [6]; the eyes, cheeks and mouth, which are very important for expression recognition, are all distorted. Their shapes and their position relationship are changed and distorted [7], [8]

Methods
Results
Conclusion

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call

Disclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.