A DEEP LEARNING APPROACH FOR DETECTING AUTISM FROM FACIAL IMAGES BLOCK DIAGRAM

In this video, a deep learning approach for detecting autism from facial image is demonstrated. The process starts with preprocessing the image using an adaptive Gabor filter, followed by data augmentation to improve the dataset. Features are extracted using MSER and selected through Gray Wolf Optimization. These features are then classified using DenseNet-121, and the final result is displayed through a Streamlit web app