Authors - Ayain John, S.Santhanalakshmi Abstract - Emotions serve as a unique psychological mechanism through which individuals communicate their subjective experiences of both their interactions with the external world and their internal state. Emotions are crucial in daily life and communication between individuals. They can be conveyed through different channels and in various forms, including facial expressions, physical movements, body posture, physical reactions, and speech patterns. The authors employed MOD_DHGN (Modified Deep Hour Glass Network) techniques to detect Autism Spectrum Disorder (ASD) in a grouped image with less sampling. The MOD_DHGN shows that it is capable of detecting emotion in images of autistic faces through augmentation and preprocessing. This study is unique because it constructed a system that uses facial emotion pictures from an extensive database, only targeting the distinct facial expressions of those affected by this condition for ASD-related emotion detection. Extensive testing revealed that the method's emotion detection accuracy was 92% in MOD_DHGN, 88% in DCNN, 72% in VGG-16, and 55% in RestNet classifiers.