Geospatial Image Mapping for Urban Navigation Systems
Led a data labeling project involving the annotation of high-resolution satellite and aerial images for mapping urban environments. Tasks included precise mapping of road networks, landmarks, and obstacles using bounding boxes and polylines to create accurate geospatial datasets. Handled a dataset of over 50,000 images, ensuring high accuracy (95%+ IoU) through iterative quality checks and collaboration with domain experts. The project supported AI models for autonomous vehicle navigation and route optimization, adhering to industry standards for data privacy and precision