A1111 StableDiffusion webUI Extension Usage and Recommendations
The Stable Diffusion web UI extension can be used to extend various functions and is recommended to improve the quality and efficiency of AI image generation.
How to Install Extensions
Installing extensions is different from installing models, but it is not that difficult if you follow the steps. Here are two ways to install “ADetailer” as an example.
Installing from the official list
adetailer in the search window and you will see a list of “ADetailer manipulations”.
adetailer
Install using GitHub repository
https://github.com/Bing-su/adetailer.git
How to disable/uninstall extensions
Disable extensions
Uninstall Extensions
stable-diffusion-webui/extensions folder in File Explorer and delete the folder of the extension you wish to uninstall.
How to Update Extensions
Introduction of Extensions
Here are some recommended extensions.
Generation control-related extensions
After Detailer
https://github.com/Bing-su/adetailer.gitIt automatically extracts faces and hands from the generated images and corrects them inpaint. Especially effective is the correction of faces in a pulled composition.
ControlNet
https://github.com/Mikubill/sd-webui-controlnet.gitVarious functions are added by installing “ControlNet”. You can add details to images and use useful functions such as the img2img extension, which generates images from rough drawings.
Aspect Ratio selector
https://github.com/LEv145/--sd-webui-ar-plusYou can control the ratio of the generated images with a single click. You can also customize it by editing resolutions.txt in stable-diffusion-webui/extensions/--sd-webui-ar-plus.
Recommended customization *Use # to customize to your liking.
Please refer to the GitHub repository if you wish to return to the default.
# SD
#SD1:1, 512, 512
SD3:2P, 768, 512
SD2:3L, 512, 768
# SDXL Square
#XL1:1, 1024, 1024
# SDXL Landscape
XL9:7L, 1152, 896
XL19:13L, 1216, 832
XL7:4L, 1344, 768
XL12:5L, 1536, 640
# SDXL Portrait
XL7:9P, 896, 1152
XL13:19P, 832, 1216
XL4:7P, 768, 1344
XL5:12P, 640, 1536Regional Prompter
Regional Prompter-Repository URLhttps://github.com/hako-mikan/sd-webui-regional-prompter.gitWith the Regional Prompter, you can control the composition of the image. For example, two people can be placed side by side and their features can be separated.
hug, 2girls, beautiful_face, standing, upper body, (ultra art illustrated style, masterpiece:1.3), ultra detailed, (medival:1.2), hill_side, blue_sky, horizon ADDCOMM
(black_hair, short_hair:1.2) ADDCOL
(red_hair, long_hair:1.2)Prompt-related extensions
Booru tag autocompletion
https://github.com/DominikDoom/a1111-sd-webui-tagcomplete.gitIt automatically displays tags used in Danbooru, a bulletin board for anime-related illustrations, as candidates
Dynamic Prompts
https://github.com/adieyal/sd-dynamic-prompts.gitThis is an extension that uses multiple prompts at random, similar to the one in “ComfyUI”. For example, if you type {smile|angry|cry}, 1girl in the prompt, it will use one of the three prompts at random. This is useful for batch processing when you want to randomly generate various expressions.
Management-related extensions
Civitai Helper
https://github.com/zixaphir/Stable-Diffusion-Webui-Civitai-Helper.gitWith “Civitai Helper,” you can easily manage your Civitai models.
Video-related extensions
With the announcement of 🔗OpenAI Sora, the quality of txt2mov has improved dramatically. The extensions introduced here are not quite as good as OpenAI Sora’s quality, but there will be extensions like OpenAI Sora in the near future.
Deforum
Deforum-Repository URLhttps://github.com/deforum-art/sd-webui-deforum.gitThis is an extension for creating AI-generated videos. It is more difficult to use than “AnimateDiff,” which will be introduced later, but it allows you to write prompts to specified keyframes and connect keyframes in a nice way, and also allows you to control the camera, such as zooming in.
AnimateDiff
https://github.com/continue-revolution/sd-webui-animatediff.git”AnimateDiff” is an extension that can generate videos more easily than “Deforum”. This one is simpler than “Deforum” and generates videos from prompts. Also, the video quality has been improved since motion module SD1.5 became V3.
Training-related extensions
DreamArtist
https://github.com/IrisRainbowNeko/DreamArtist-sd-webui-extension.gitThis is an extension that creates training data from a single piece of training data and can be used as a prompt. A PC with reasonable specs is required for training data.
Conclusion
In this article, we have introduced how to use and recommend extensions for Stable Diffusion web UI. New extensions are being released all the time. We would like to write another article on how to use the extensions introduced in another article.
