Image Classification and OCR data annotator
Classified and tagged large volumes of images containing text as part of an OCR (Optical Character Recognition) model training dataset, applying consistent labeling schemas to ensure high-quality ground truth data for AI development. Evaluated image quality attributes including resolution, legibility, and text clarity, flagging unsuitable samples to maintain dataset integrity. Applied detailed annotation guidelines to categorize images by text type, layout, language, and contextual relevance across diverse real-world sources such as street signs, documents, receipts, and printed materials. Maintained consistently high accuracy scores on internal quality checks, meeting and exceeding Appen's inter-annotator agreement benchmarks throughout the engagement. Processed high daily volumes of images while sustaining labeling precision, demonstrating the ability to balance speed and quality in a production annotation environment. Adapted quickly to updated labeling instructions and schema revisions across project iterations, ensuring continuity and consistency in output quality.