Autism Detection From Image Using Generative AI

Abstract

Autism Spectrum Disorder (ASD) Is A Neurodevelopmental Condition Characterized By Challenges In Social Interaction, Communication, And Behaviour. Early Detection Plays A Crucial Role In Improving Intervention Outcomes, Yet Traditional Diagnostic Methods Are Often Time-Consuming, Subjective, And Require Expert Evaluation. This Study Proposes An Innovative Approach For Autism Detection Using Image-Based Analysis Powered By Generative Artificial Intelligence Techniques. The Proposed System Leverages Deep Learning And Generative Models, Such As Generative Adversarial Networks (Gans) And Diffusion Models, To Analyse Facial Features, Eye Gaze Patterns, And Behavioural Cues Captured In Images. Generative AI Is Employed Not Only To Enhance Data Quality Through Synthetic Data Augmentation But Also To Improve Model Robustness In Scenarios With Limited Labelled Datasets. A Convolutional Neural Network (CNN) Is Integrated With Generative Models To Extract Discriminative Features And Classify Individuals As Autistic Or Non-Autistic. Experimental Results Demonstrate That The Integration Of Generative AI Significantly Improves Classification Accuracy, Reduces Overfitting, And Enhances Generalization Across Diverse Datasets. The System Offers A Non-Invasive, Cost-Effective, And Scalable Solution For Preliminary Autism Screening. However, Ethical Considerations, Including Data Privacy, Bias Mitigation, And Clinical Validation, Are Critical For Real-World Deployment. This Research Highlights The Potential Of Generative AI In Transforming Early Autism Detection And Supports The Development Of Assistive Diagnostic Tools For Healthcare Professionals.