Deep Learning and IoT Based Assistance System for Autism Spectrum Disorder People

During the Covid-19 pandemic situation, the economic and social disruption is devastating. People could not get out of their homes and lead their normal life. Schools across the country have switched to remote learning, which is also inevitable. Though there are some advantages in the online classes, the fact that many of the children suffered mental stress due to these classes could not be denied. Especially, when speaking about the autism people, they could not handle the stress like normal children. The caretaker is necessary to assist them all the time. Therefore, an assistance system is needed to monitor the stress level. In this research, prototype of deep learning and Internet of Things (IoT) based assistance system has been proposed. It monitors the stress parameters namely body temperature, pulse rate, skin conductivity and the facial emotion of the autism disorder people. Further, the hardware model (Raspberry Pi) has been developed to measure the stress level. The processed data from the model has been stored in the Thingspeak cloud platform for monitoring the autism people remotely. From the threshold stress parameters, the level of the stress can be predicted by the proposed stress management algorithm.