Professional-grade image processing functions in detail
In addition to basic text to generate images, the platform provides four major advanced functions to meet professional needs:
- Upscale optimization: Select a certain version (numbered 1-4) of the initially generated image for detail enhancement, call example:
curl -X POST /upscale -d '{"hash":"[task id]", "choice" :2}' - Vary Region: Specify the coordinate area on the generated image (need to be Upscale first), with the new cue word local redrawing, applicable to the product to change the background and other scenarios.
- Creative Derivative Model (Remix): When the /prefer interface is turned on, the cue word can be iterated while maintaining the style of the original drawing, which is particularly suitable for the evolution of design concepts.
- Reverse Description Generation (Describe): Upload images to get 4 text descriptions to provide inspirational direction for subsequent revisions
Operation points: All editing functions need to be based on the initial generated image task ID to start, so it is recommended to use /imagine to create a base image first. Region editing requires knowledge of Discord's coordinate system (0-1 standardized coordinates), which is illustrated in detail in the documentation. For e-commerce users, with Upscale + Vary Region can realize the rapid batch transformation of product images.
This answer comes from the articleUserAPI.ai: Integration of Midjourney image generation API interfaceThe
May not be reproduced without permission:AI productivity tools " What advanced image editing features does UserAPI.ai support? How do I use them?
