https://tinyurl.com/bdhmpw66Designed and implemented high-level ResNet and UNet models to segment buildings using remote-sensing data with high accuracy. It achieved 86% accuracy, which significantly enhanced the capacity of spatial analysis in remote-sensing applications. Engineered and optimized the deep learning model for using YOLO on satellite image object detection. Improved the accuracy to over 70%, improving the capability of object detection and classification in satellite imagery. Design and implement a segmentation model that detects and blurs Personally Identifiable Information with accuracy above 72%.