Facial expression recognition and generation using sparse autoencoder

Release time:2021-12-25  Hits:

• DOI number:10.1109/SMARTCOMP.2014.7043849
• Journal:2014 International Conference on Smart Computing (SMARTCOMP)
• Key Words:Decoding, Image reconstruction, Vectors, Integrated circuits, Approximation methods
• Abstract:Facial expression recognition has important practical applications. In this paper, we propose a method based on the combination of optical flow and a deep neural network - stacked sparse autoencoder (SAE). This method classifies facial expressions into six categories (i.e. happiness, sadness, anger, fear, disgust and surprise). In order to extract the representation of facial expressions, we choose the optical flow method because it could analyze video image sequences effectively and reduce the influence of personal appearance difference on facial expression recognition. Then, we train the stacked SAE with the optical flow field as the input to extract high-level features. To achieve classification, we apply a softmax classifier on the top layer of the stacked SAE. This method is applied to the Extended Cohn-Kanade Dataset (CK+). The expression classification result shows that the SAE performances the classification effectively and successfully. Further experiments (transformation and purification) are carried out to illustrate the application of the feature extraction and input reconstruction ability of SAE.
• Co-author:Xueshi Hou,Jiansheng CHEN, Chang Yang, Guangda Su, Weibei Dou,Weibei Dou, Yinyang Wang,Yinyang Wang, Tao Jiang, Jean-Marc Constans
• First Author:Yunfan Liu,Xue Wang
• Indexed by:会议论文
• Correspondence Author:Jiansheng CHEN,Weibei Dou
• Page Number:125-130
• Translation or Not:no
• Date of Publication:2014-11-03