Emotion Detection using Bio Signals and usefulness for Autism Spectrum Disorder

Autism Spectrum Disorder (ASD) which is neurological and developmental condition. ASD can be detected by observing distinct patterns in an individual's learning, behaviour, communication, decision making ability and social interactions. The ability to anticipate emotions can prove beneficial in various situations, particularly when dealing with individuals across the autism spectrum with varying degrees of severity. Decision- making encompasses not only intelligence and perception but also emotions. Emotion Detection is helpful in Bio Feedback Therapy for ASD. Positive emotions, for example mostly tend to facilitate forward-thinking and successful decision-making while negative emotions often contribute to poor well-being or depression. A transformer model, a type of neural network architecture, possesses the capability to automatically convert one form of input into another form of output. This work intends to investigate various signals that may be useful in identifying emotions for the purpose of predicting autism as well as transformer model-based algorithms that may be useful in determining the Autism Severity Grading. Similar methods are used now to assist those who are autistic. There is also discussion of medical applications using minimal parameters and methods.