Dataset Creator and Image Annotator for Metro Monitoring System
Created a dataset of 5,000 annotated images for a computer vision project targeting metro seat occupancy detection. Used Roboflow and manual labeling to delineate seat regions and train object detection algorithms. Ensured annotation quality for YOLO-based model training and evaluation. • Labeled images with clear seat boundaries using bounding box techniques. • Utilized Roboflow for annotation management and dataset versioning. • Performed manual quality checks to reduce noise and mislabels. • Supported automated system design for seat occupancy alerts.