Developed a segmentation framework to use Contextual Hierarchical Models with CNNs to segment images and produced results with improved pixel-wise and class average accuracy. - View it on GitHub