Autism Spectrum DisorderAutism spectrum disorder (ASD) is a neurodevelopmental condition affecting early childhood and impacting diverse aspects of a child’s development. In the absence of medical tests, the observation of behavioral patterns remains the primary diagnostic approach for ASD. Recently, neural network-based models have emerged for early ASD identification. In this paper, we present an ASD predictionPrediction model utilizing Convolutional Neural NetworkConvolutional Neural Network (CNN) architectures, specifically EfficientNet-B7, in combination with Haar Cascade. This approach offers a significant improvement in the accuracy and predictionPrediction of ASD in children. This model is integrated into a web application for image-based ASD predictionPrediction using Dash module from Python. This application helps healthcareHealthcare professionals with a practical tool for early ASD identification.