AI Data Trainer — Image Generation & Diffusion Models
As an AI Data Trainer specializing in image generation and diffusion models, I evaluated, annotated, and provided quality feedback on AI-generated images. My work involved assessing prompt accuracy, visual coherence, and adherence to style and quality criteria in outputs from tools such as Stable Diffusion, Midjourney, and DALL·E. I consistently labeled compositional and stylistic properties to enhance model performance for generative AI applications. • Assessed AI-generated images for prompt fidelity, visual coherence, and artifact detection • Labeled and categorized images by object, environment, lighting, and composition • Leveraged generative tools and CLIP-based scoring for data evaluation workflows • Maintained high-throughput annotation standards for large datasets