Skip to content

Unofficial implementation of "SEGA: Instructing Text-to-Image Models using Semantic Guidance". Semantic Guidance gives you more control over the semantics of an image given an additional text prompt. An extension for Automatic1111 WebUI.

License

Notifications You must be signed in to change notification settings

v0xie/sd-webui-semantic-guidance

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

55 Commits
 
 
 
 
 
 
 
 

Repository files navigation

sd-webui-semantic-guidance

An unofficial implementation of SEGA: Instructing Text-to-Image Models using Semantic Guidance for Automatic1111 WebUI

Semantic Guidance gives you more control over the fine details of an image by emphasizing or de-emphasizing concepts given in an additional positive / negative text prompt.

Use Cases

  • Altering the style of an image without changing the composition image

  • Adding concepts to an image image

Please see the bottom of the page for more examples.

Issues / PR's are welcome!

Settings

  • Positive / Negative Prompt: Concepts to emphasize / de-emphasize, separated by commas
  • Warmup Steps: How many steps to wait before applying semantic guidance
  • Edit Guidance Scale: Globally scale how much influence semantic guidance has on the image
  • Tail Percentage Threshold: The percentage of latents to use when calculating the semantic guidance
  • Momentum Scale: Scale the influence of the added momentum term
  • Momentum Beta: Higher values will make the influence of the momentum term more stable

Feature / To-do List

  • SD XL support
  • Support A1111 prompt attention syntax and shortcuts for attention strength
  • Support restoring parameter values from infotext (Send to Txt2Img, Send to Img2Img, etc.)
  • XYZ plot support
  • SD 1.5 support (untested)
  • Support semantic guidance prompt in regular Positive/Negative Prompt fields

Credits

More examples

  • Enhancing the details of an image image

About

Unofficial implementation of "SEGA: Instructing Text-to-Image Models using Semantic Guidance". Semantic Guidance gives you more control over the semantics of an image given an additional text prompt. An extension for Automatic1111 WebUI.

Topics

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages