Text-to-image
Text-to-image: Generating images guided by a textual description. A text encoder conditions the visual model, which samples an image consistent with the prompt. Composition, text and physical constraints remain prone to errors and bias. Examples include Prompt-based illustration, Concept art, Campaign variations.
In italiano: Text-to-ImageLegge II · Probabilità, softmax e modelli generativiA text encoder conditions the visual model, which samples an image consistent with the prompt. Composition, text and physical constraints remain prone to errors and bias.
Examples
- Prompt-based illustration
- Concept art
- Campaign variations