AI Data Annotator & Image Model Trainer (visual evaluation and descriptive annotation)
ImageAssessed and evaluated visual content for multimodal image verification tasks using detailed descriptive text. Translated visual properties (lighting, composition, color, and setting) into structured natural-language descriptions to support model training and output quality improvement. Repeatedly applied descriptive precision to provide consistent, high-signal annotations or ratings for AI systems. • Evaluated image characteristics such as contrast, angle, reflections, and lighting sources. • Wrote nuanced, domain-relevant scene descriptions aligned to target verification goals. • Used observation-to-text translation to improve training/evaluation consistency. • Applied aesthetic and structural critique (balance and hierarchy) to guide model optimization.