Computer Vision Data Annotation & Object Detection
This project involved creating high-precision ground truth datasets for computer vision models. I performed precise object localization and identification using tools like [mention a tool if you know one, e.g., CVAT, Labelbox, or RectLabel]. Tasks included drawing accurate bounding boxes and complex polygons to delineate objects, identifying key landmarks for pose estimation, and performing semantic segmentation to distinguish foreground objects from backgrounds. I ensured that all annotations met strict pixel-perfect quality standards to improve model accuracy in object recognition tasks.