AdvancedEAST is a computer vision algorithm for detecting text regions in scene images. It builds upon the EAST (Efficient and Accurate Scene Text Detector) architecture with improvements to handle long text predictions more accurately. Implemented in Keras with TensorFlow, the project includes network architecture definitions, custom loss functions, data preprocessing, training pipelines, and post-processing utilities for extracting text bounding box coordinates from images.