AI Training Data Specialist – Self-Directed Project (Image annotation, quality evaluation, and training support)
Conducted quality evaluation and annotation of 300+ AI-generated images for accuracy and visual coherence. Assessed adherence to prompt specifications and identified errors in model outputs using structured criteria. Maintained consistency through iterative calibration and reference cross-checking across multiple rounds of labeling. • Labeled images using a quality rubric (excellent/good/acceptable/unusable) • Evaluated anatomical errors, perspective issues, and style inconsistencies • Ensured coherence between prompt requirements and generated visual content • Produced feedback-ready notes reflecting observed issues and deviations