基于协方差池的面部表情识别及其教学监督应用

周 子涵, 李 丹
四川大学锦城学院

摘要

二十一世纪以来,伴随着人们生活质量的提高,信息传递的方式也是不断的改革变化。其中,图像传递尤为重要。因此，人脸识别、面部表情识别等技术的运用领域也越来越广。将人脸表情进行分类需要抓取脸部的一些关键点。但是传统的CNN对面部关键点的捕获效果一般，那么如何更好地获得面部特征中的关键点呢？本文将研究协方差这种二阶统计量[1]在抓取面部关键点的优势。因此，本文主要介绍了基于CK+和SFEW2.0数据集下的
协方差池面部表情识别系统，并将其应用到了教学领域，进行教学监督。

关键词

表情识别；协方差；教学监督

参考

[1] K. Yu and M. Salzmann. Second-order convolutional neural net- works. CoRR, abs/1703.06817, 2017. 1, 2, 3, 7

[2] B.-K.Kim,H.Lee,J.Roh,and S.-Y.Lee.Hierarchicalcommitteeof- deepcnnswithexponentially-weighteddecisionfusion for static facial expression recognition. In Proceedings ofthe2015ACMon- InternationalConferenceonMultimodal Interaction,ICMI’15,pag- es427–434,NewYork,NY,USA, 2015. ACM. 2, 5

[3] Z. Yu and C. Zhang. Image based static facial expression recog- nition with multiple deep network learning. In Proceedings of the

2015 ACM on International Conference on Multimodal Interac- tion, ICMI ’15, pages 435–442, New York, NY, USA, 2015. ACM. 2, 5

[4] J. a. Carreira, R. Caseiro, J. Batista, and C. Sminchisescu. Se- mantic segmentation with second-order pooling. In Proceedings of the 12th European Conference on Computer Vision - Volume Part VII, ECCV’12, pages 430–443, Berlin, Heidelberg, 2012. Springer-Verlag. 2, 3

[5] O. Tuzel, F. Porikli, and P. Meer. Region covariance: A fast de- scriptor for detection and classi-cation. In Proceedings of the 9th European Conference on Computer Vision -Volume PartII,EC- CV’06,pages589–600,Berlin,Heidelberg,2006. Springer-Ver- lag. 1, 2, 3

[6] K. Yu and M. Salzmann. Second-order convolutional neural net- works. CoRR, abs/1703.06817, 2017. 1, 2, 3, 7

[7] Z. Huang and L. V. Gool. A riemannian network for spd matrix learning. In AAAI, 2017. 1, 2, 3, 7

[8] 许春燕 . 基于黎曼流形的图像分类算法研究 [D]. 武汉 : 华中 科技大学 ,2015.

[9] J. a. Carreira, R. Caseiro, J. Batista, and C. Sminchisescu. Se- mantic segmentation with second-order pooling. In Proceedings of the 12th European Conference on Computer Vision - Volume Part VII, ECCV’12, pages 430–443, Berlin, Heidelberg, 2012. Springer-Verlag. 2, 3

[10] O. Tuzel, F. Porikli, and P. Meer. Region covariance: A fast de- scriptor for detection and classi-cation. In Proceedings of the 9th European Conference on Computer Vision - Volume PartII,EC- CV’06,pages589–600,Berlin,Heidelberg,2006. Springer-Ver- lag. 1, 2, 3

[11] Z. Huang and L. V. Gool. A riemannian network for spd matrix learning. In AAAI, 2017. 1, 2, 3, 7

[12]Sophus Lie’s Transformation Groups: A Series of Elementary Articles. 1897, 4(12):308-313.

[13] X. Zhu and D. Ramanan. Face detection, pose estimation and- landmarkestimationinthewild. InIEEEConferenceon Computer Vision and Pattern Recognition (CVPR), 2012. 4

[14]薛同来,赵冬晖,张华方,郭玉,刘旭春.基于Python的深 度学习人脸识别方法 [J]. 工业控制计算机 ,2019,32(02):118- 119.

[15] 林志健 , 周设营 , 陈延清 . 基于 OpenCV 的人脸识别关键技 术分析 [J]. 中国新技术新产品 ,2020(07):15-16.

[16] 李宽 . 基于浅层卷积网络的人脸表情识别方法研究 [D]. 合 肥 : 中国科学技术大学 ,2019.