Emotion Recognition Techniques with IoT and Deep Learning Technologies

Abstract

The emerging technologies of the so-called fourth industrial revolution (4IR) are part of many processes carried out by people in the contemporary era, where different applications can be found in the field of the Internet of Things (IoT), Big Data, Machine and Deep Learning, among others. On the other hand, a situation that is currently gaining relevance, after the Covid-19 pandemic, is the understanding and analysis of people’s emotions, supported by techniques that involve IoT and Deep Learning technologies, such as the proposal in this document. The work carried out seeks to introduce the reader to the concept and techniques of facial recognition and emotions, as an alternative in the medium term that allows the development of a low-cost platform, which will be based on the use of some validated algorithms and processes, used and developed for this field. Likewise, as a complement and part of the platform under development, the use of IoT devices in it is studied, which will help in the task of recognizing emotions in people. Therefore, in this document, apart from the software used, some IoT hardware solutions are explained, which have been tested for the aforementioned objective. Finally, the document describes an application case made with the mentioned algorithms and the proposed hardware, the work is the beginning of a more complex investigation that in this document presents its preliminary results.