Towards the Edge Intelligence: Robot Assistant for the Detection and Classification of Human Emotions

Abstract

Deep learning is being introduced more and more in our society. Nowadays, there are very few applications that do not use deep learning as a classification tool. One of the main application areas is focused on improving people's life quality, allowing to create personal assistants with canned benefits. More recently, with the proliferation of mobile computing and the emergence of the Internet of Things (IoT), billions of mobile and IoT devices are connected to the Internet. This allows the generation of millions of bytes of information about sensors, images, sounds, etc . Driven by this trend, there is an urgent need to push the IoT frontiers to the edge of the network, in order to decrease this massive sending of information to large exchanges for analysis. As a result of this trend, a new discipline has emerged: edge intelligence or edge AI, a widely recognised and promising solution that attracts with special interest to the community of researchers in artificial intelligence. We adapted edge AI to classify human emotions. Results show how edge AIbased emotion classification can greatly benefit in the field of cognitive assistants for the elderly or people living alone.

Artificial Intelligence Trends and Tools for Improving Women’s Health

“…AI technologies play a crucial role in addressing mental health issues among women by offering personalized interventions, remote monitoring solutions, and mental health assessment tools. Chatbots, virtual assistants, and AI-based therapy platforms provide accessible and stigma-free support for women experiencing mental health challenges, promoting early intervention, and improving overall wellbeing [59].…”

A Systematic Literature Review on Distributed Machine Learning in Edge Computing

“…Rincon et al [ 117 ] present an assistant robot based on EI, which incorporates two devices capable of classifying emotional states and physical activities performed by the user. The training uses Keras [ 118 ], and after obtaining the H5 model, there is a transformation to TensorFlow lite and, finally, the K model is obtained with a Mobilenet network [ 119 ].…”

“…The recent effort in integrating IoT into the healthcare scenario has been a great catalyst to the development of novel smart care applications or the enhancement of past ones, which may include telemedicine, tele-consultancy, pervasive health monitoring applications, human activity [ 101 ] and/or emotion recognition [ 117 ], and others.…”