Data Annotator Remotask
Performed high-precision image annotation for a computer vision dataset. Tasks included drawing tight bounding boxes and detailed polygon masks around various objects such as , vehicles, pedestrians, and traffic signs across 12,400 diverse real-world images. Ensured pixel-level accuracy while maintaining an average IoU score above 0.92. Strictly followed complex annotation protocols, including hierarchical labeling and attribute tagging. Contributed to dataset quality through multiple review cycles, resulting in production-ready training data.