Research Article
Six-layer Optimized Convolutional Neural Network for Lip Language Identification
@ARTICLE{10.4108/eai.20-8-2021.170751, author={Yifei Qiao and Hongli Chen and Xi Huang and Juan Lei and Xiangyu Cheng and Huibao Huang and Jinghan Wu and Xianwei Jiang}, title={Six-layer Optimized Convolutional Neural Network for Lip Language Identification}, journal={EAI Endorsed Transactions on e-Learning}, volume={7}, number={22}, publisher={EAI}, journal_a={EL}, year={2021}, month={8}, keywords={lip language identification, convolutional neural network, Batch Normalization, dropout}, doi={10.4108/eai.20-8-2021.170751} }- Yifei Qiao
Hongli Chen
Xi Huang
Juan Lei
Xiangyu Cheng
Huibao Huang
Jinghan Wu
Xianwei Jiang
Year: 2021
Six-layer Optimized Convolutional Neural Network for Lip Language Identification
EL
EAI
DOI: 10.4108/eai.20-8-2021.170751
Abstract
INTRODUCTION: Lip language is one of the most important communication methods in social life for people with hearing impairment and impaired expression ability. This communication method relies on visual recognition to understand the meaning expressed in communication.
OBJECTIVES: In order to improve the accuracy of this natural language recognition, we propose six-layer optimized convolutional neural network for lip recognition.
METHODS: The calculation method of the convolutional layer in the CNN model is used, and two pooling methods are compared: the maximum pooling operation and the average pooling operation to analyse the most important feature data in the picture. In order to reduce the simulation in the model training process, the closing rate has been optimized by introducing Dropout technology.
RESULTS: It shows that the recognition accuracy rate based on the six-layer convolutional neural network can reach 85.74% on average. This method can effectively recognize lip language.
CONCLUSION: We propose a six-layer optimized convolutional neural network method for lip language recognition, and the identification of lip language features of this method is better than 3D+ DenseNet +1 × 1 Conv +resBi-LSTM, 3D+CNN, ConvNet+2 -256-LSTM+VGG-16 three advanced methods.
Copyright © 2021 Yifei Qiao et al., licensed to EAI. This is an open access article distributed under the terms of the Creative Commons Attribution license, which permits unlimited use, distribution and reproduction in any medium so long as the original work is properly cited.