ComfyUI Node: Text-to-Image Prompt Enhancer

Text-to-Image Prompt Enhancer:

The EricTextToImagePromptEnhancer is a sophisticated tool designed to refine and enhance text prompts for generating images using AI models. Its primary purpose is to take a basic text description and enrich it with additional context, keywords, and platform-specific requirements to produce a more detailed and effective prompt for image generation. This node leverages local language models to enhance the prompt, ensuring that the resulting image is more aligned with the user's creative vision. By processing alternations, preserving emphasis syntax, and incorporating reference images, the node provides a comprehensive enhancement process that boosts the density and specificity of the prompt. This results in more accurate and visually appealing image outputs, making it an invaluable tool for AI artists looking to optimize their creative workflows.

Text-to-Image Prompt Enhancer Input Parameters:

text_prompt

The text_prompt parameter is a string input that serves as the initial description of the image you want to generate. It supports multiline input and allows for advanced syntax such as emphasis, de-emphasis, and alternation to fine-tune the weight and selection of keywords. The default value is "a beautiful woman in a garden," and it acts as the foundation upon which the node builds and enhances the prompt.

reference_directive_1

The reference_directive_1 parameter provides guidance on how to interpret the first reference image, if provided. It offers several options, including "none," "auto," "recreate," "reinterpret," "subject only," "style only," "lighting only," "composition," and "genre," with "none" as the default. This parameter influences how the reference image is used to enhance the text prompt, allowing for specific aspects of the image to be emphasized or altered.

reference_directive_2

Similar to reference_directive_1, the reference_directive_2 parameter applies to a second reference image. It shares the same options and default value, providing additional flexibility in how multiple reference images can be utilized to enrich the text prompt.

prompt_context

The prompt_context parameter determines the overall approach to enhancing the text prompt. Options include "none," "auto," "expand_short_prompt," "finish_opening_line," "prompt_from_item_list," "modify_reference_image," and "enhance_reference_image," with "none" as the default. This parameter guides the node in selecting the most appropriate enhancement strategy based on the user's creative goals.

Text-to-Image Prompt Enhancer Output Parameters:

positive_prompt

The positive_prompt output is a string that represents the enhanced version of the original text prompt. It incorporates additional keywords, context, and platform-specific requirements to create a more detailed and effective prompt for image generation.

negative_prompt

The negative_prompt output is a string that identifies elements or features to avoid in the generated image. This helps refine the image generation process by specifying what should not be included, ensuring a more focused and accurate result.

image_description

The image_description output provides a textual summary of the expected image based on the enhanced prompt. It serves as a guide for understanding the intended visual outcome of the image generation process.

status

The status output indicates the success or failure of the prompt enhancement process. It provides feedback on whether the local language model successfully enhanced the prompt or if a fallback method was used.

Text-to-Image Prompt Enhancer Usage Tips:

• Utilize the text_prompt parameter's advanced syntax to fine-tune the emphasis and selection of keywords, enhancing the specificity of your image description.
• Experiment with different reference_directive options to see how various aspects of reference images can influence the enhanced prompt and resulting image.

Text-to-Image Prompt Enhancer Common Errors and Solutions:

LLM failed or returned empty content - using deterministic fallback

• Explanation: This error occurs when the local language model is unable to enhance the prompt, possibly due to an issue with the model or the input text.
• Solution: Check the input text_prompt for any errors or unsupported syntax. Consider simplifying the prompt or adjusting the prompt_context to better align with the model's capabilities.