An Intelligent Facial Expression Recognizer Using Modified ResNet-110 Using Edge Computing

Abstract

In this intelligent society, facial expression recognition is an important way to make assumptions about people mood and mentation. It can be applied in security and risk analysis. Many edge computing applications need facial expression recognition as the basic support. Some shortcomings have been observed in the narrative probe, such as dirty universalization ability of network architecture and low firmness of recognition system. This chapter proposes an intelligent computing method based on ResNet-110. We entered this model into the experiment and achieved a 96.29% accuracy. The laboratory findings imply that the proposed is outperform the main current recognition models. In order to express our thoughts clearly, the method can be applied to the edge computing environment.