Performance Analysis of Emotion Recognition Prediction on Mobile Devices

Artificial Intelligence (AI) has had a boom in recent years thanks to technological development, it is increasingly used in various fields of research and many of these use mobile devices. Emotion recognition is one of them, as it helps in neuroscience, computer science, and medical applications. There is information where applications that are developed on mobile devices for emotion recognition, but none considers the performance of the algorithm used. The purpose of this article was to evaluate how it affects the performance of a mobile application based on Artificial Intelligence (AI) for emotion recognition according to the properties of mobile devices of different ranges. For this purpose, two different Convolutional Neural Network (CNN) architectures are evaluated, which will be analyzed according to the following metrics: response time, RAM, CPU usage, and accuracy. The results show that a deep layered CNN has better performance and lower computational cost compared to a conventional CNN on mobile devices.