Research on the Blind Auxiliary Communication System Based on Deep Learning
In order to solve the communication obstacles caused by the visual defects of the blind,combining the research on ordinary people's communication theory with the existing model of communication,the blind communication model and the blind auxiliary communication model are constructed.Through the combination of hardware and software,the blind auxiliary communication system is designed.The hardware system mainly includes two parts of smart glasses and portable computing devices.The software system mainly collects images and sounds,identifies communication targets and their external features,and expression communication behaviors.Among them,expression recognition technology uses ResNet18 Residual Neural Network to achieve real-time expression recognition function,and finally the sound is synthesized and the result is output.
the blindAC modelauxiliary communicationexpression recognition
国家科技期刊平台
NSTL
万方数据
