What are the limitations of text-to-image generation when creating photorealistic human figures?

Text-to-image generation, while powerful, faces several limitations when creating photorealistic human figures. One major challenge is accurately generating realistic facial features and expressions. Human faces are incredibly complex, and subtle variations in shape, texture, and lighting can significantly impact perceived realism. AI models often struggle to capture these nuances, leading to faces that look unnatural or uncanny. Generating realistic skin textures is another significant hurdle. Human skin has a complex microstructure and su....