SD Pilot-Stable Diffusion prompt optimization

Example

A user submits a prompt like 'a robot in a cityscape.' SD Pilot analyzes the prompt to identify the key elements.

Scenario

By breaking down 'a robot in a cityscape' into components such as [futuristic style], [robot], [cityscape], [metallic texture, modern buildings], [high-tech feel], [standing pose], [mid-shot], [futuristic city with neon lights], [bright artificial lighting], [low-angle shot], [photorealistic], SD Pilot transforms a vague prompt into a detailed and structured one.

Example

Enhancing a simple prompt like 'a cat on a chair' to a more detailed one.

Scenario

The prompt 'a cat on a chair' becomes '[realistic photo], [a fluffy cat], [wooden chair], [calm and relaxed], [cat sitting], [portrait shot], [cozy living room], [soft daylight], [eye-level angle], [digital photography].' This level of detail helps in creating a precise and appealing image.

Example

Creating a prompt from scratch based on user requirements.

Scenario

A user wants an image of 'a medieval knight in battle.' SD Pilot constructs the prompt as '[dramatic oil painting], [medieval knight], [battle scene], [armor, sword, shield], [intense and dynamic], [fighting pose], [wide shot], [battlefield with other knights], [dramatic lighting], [low-angle shot], [style of Frank Frazetta].' This custom prompt aligns with the user's vision and artistic intent.