Image Annotation Specialist (AI Training Data)
As an Image Annotation Specialist at Mercor, I refined and reviewed bounding boxes, object labels, and captions over large-scale image datasets for AI training. My work involved identifying incorrect labels, adjusting bounding boxes, and matching objects across paired images to ensure dataset accuracy and consistency. I maintained a high quality of annotated data by following structured guidelines, supporting improved machine learning performance. • Reviewed and refined bounding boxes for spatial accuracy and consistency. • Added missing bounding boxes and classified visual elements including logos, text segments, and common objects. • Corrected transcription errors and captions for improved label accuracy. • Verified annotation quality based on correctness and adherence to project guidelines.