Facial Expression Recognition Based on Deep Binary Convolutional Network

被引：0
|
作者：
Zhou L. [1 ,2 ,3 ]
Liu J. [1 ,2 ]
Li W. [2 ]
Mi J. [2 ]
Lei B. [3 ]
机构：
[1] College of Software, Chongqing University of Posts and Telecommunications, Chongqing
[2] Key Laboratory of Image Cognition, Chongqing University of Posts and Telecommunications, Chongqing
[3] Hubei Key Laboratory of Intelligent Vision Based Monitoring for Hydroelectric Engineering, China Three Gorges University, Yichang
来源：
Jisuanji Fuzhu Sheji Yu Tuxingxue Xuebao/Journal of Computer-Aided Design and Computer Graphics | 2022年 / 34卷 / 03期
关键词：
Attention mechanism; Binary convolutional network; Facial expression recognition; Local binary pattern;
D O I：
10.3724/SP.J.1089.2022.18920
中图分类号：
学科分类号：
摘要：
Facial expression recognition (FER) is a challenging task because of the large number of parameters, low speed and insufficient feature representation of the expression region. In order to address the above-mentioned challenges, a deep binary convolutional network for FER is proposed. Firstly, a lightweight convolution network (BRNet) with parallel operation of binary convolution and traditional convolution is designed to reduce the complexity of network and improve the speed of recognition. Secondly, a dynamic radius strategy is constructed to extract binary features and form the attention weight of expression region so that the local features and global features can be fused effectively. Finally, the cross-entropy and L2 loss are designed for quick and accurate expression classification. Experiments show that the average accuracy of the proposed method is 99.25% and 93.85% on CK+ and Oulu-CASIA respectively, which are higher than the other similar lightweight convolutional networks. The parameters and computation cost of network are 5.0×105 bytes and 2.1×106 bytes. In contrast, the computation cost of EfficientFace is about 77 times than proposed method. As a result, the effectiveness and lightness of proposed method have been proved. © 2022, Beijing China Science Journal Publishing Co. Ltd. All right reserved.
引用
收藏
页码：425 / 436
页数：11
相关论文
共 43 条
  • [21] Zou W, Zhang D, Lee D J., A new multi-feature fusion based convolutional neural network for facial expression recognition, Applied Intelligence, pp. 1-12, (2021)
  • [22] Zhao Z Q, Liu Q S, Zhou F., Robust lightweight facial expression recognition network with label distribution training, Proceedings of the 35th AAAI Conference on Artificial Intelligence, pp. 3510-3519, (2021)
  • [23] Qin H T, Gong R H, Liu X L, Et al., Binary neural networks: a survey, Pattern Recognition, 105, (2020)
  • [24] Rastegari M, Ordonez V, Redmon J, Et al., XNOR-Net: ImageNet classification using binary convolutional neural networks, Proceedings of European Conference on Computer Vision, pp. 525-542, (2016)
  • [25] Xu F, Boddeti V N, Savvides M., Local binary convolutional neural networks, Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, pp. 4284-4293, (2017)
  • [26] Zhu S L, Dong X, Su H., Binary ensemble neural network: more bits per network or more networks per bit?, Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pp. 4918-4927, (2019)
  • [27] Chen Yong, Liu Xi, Liu Huanlin, Occluded pedestrian detection based on joint attention mechanism of channel-wise and spatial information, Journal of Electronics & Information Technology, 42, 6, pp. 1486-1493, (2020)
  • [28] Vincent O, Folorunso O., A descriptive algorithm for Sobel image edge detection, Proceedings of the InSITE Conference, pp. 97-107, (2009)
  • [29] Lucey P, Cohn J F, Kanade T, Et al., The extended Cohn- Kanade dataset (CK+): a complete dataset for action unit and emotion specified expression, Proceedings of the IEEE Computer Society Conference on Computer Vision and Pattern Recognition Workshops, pp. 94-101, (2010)
  • [30] Goodfellow I J, Erhan D, Carrier P L, Et al., Challenges in representation learning: a report on three machine learning contests, Proceedings of International Conference on Neural Information Processing, pp. 117-124, (2013)
12345