Deep training with structured neural models has emerged as a dominant paradigm in synthetic intelligence. These systems are crafted to process information by recognizing features. Convolutional layers allow for the optimized representation of numerical information, making them particularly suitable for applications such as video segmentation. Ex