Photoshop AI generation tools other than Firefly Auto Photoshop Stable Diffusion plugin

A few days ago, Adobe opened Firefly for public testing, as long as you have an Adobe ID, you can use it. I believe that many Photoshop users have tried its ability to generate images in the past few days. Firefly emphasizes that the generated images have no copyright and security issues, but if you also want to have the freedom of Stable Diffusion to generate beautiful pictures, a recent Photoshop plug-in can realize your wish, and the Stable Diffusion WebUI (below called WebUI) combined with the image editing capabilities of Photoshop.

The Auto Photoshop Stable Diffusion plug-in calls its generation function through the API function of WebUI, so the premise of using it is to install the WebUI of AUTOMATIC1111 on the same computer.For installation and use of AUTOMATIC1111 WebUI, you can refer tothis articleteaching.

The image is yours! Self-built WebUI image generation platform:click here
Auto Photoshop StableDiffusion plugin GitHub page:click here

The Auto Photoshop Stable Diffusion plug-in can use most of the functions of WebUI and supports all checkpoint models and LoRA, but the support for extended functions is limited. fortunately! It supports the most important extension, Controlnet!

Two RAM-eating monsters walking together

In terms of system requirements, the minimum version of Photoshop required by the plug-in is 24, but because both Photoshop and Stable Diffusion are programs that earn resources, even if the VRAM of the display card has 16GB, the system memory should have 32GB or more, and when it is found that “CUDA out of memory ” problem, consider adding in the WebUI configuration file --medvram parameter.

Installation: WebUI

When using the Auto Photoshop Stable Diffusion plug-in, you need to start the WebUI at the same time, and you need to modify the file of the startup program webui-user.bat,exist set COMMANDLINE_ARGS= one entry --api Parameters to enable the API program interface function.

1. Enter the folder of WebUI, right click webui-user.bat file and select “Edit” in the pop-up menu;

2. exist set COMMANDLINE_ARGS= one entry --api Parameters, the original parameters should be retained. Save and close the file.

However, the official also provides an Auto-Photoshop-StableDiffusion extension component of the same name to WebUI in order to use the plug-in’s smart mask and image search functions.

1. in the plugin’s GitHub page Click the green “Code” button and copy the “HTTPS” link address of “Clone”;

2. Paste the URL into the “Install from URL” on the WebUI Extensions page, and then click the “Install” button to start the installation;

3. During the installation process, you can see the progress in the command line window of WebUI;

4. After the installation is complete, click “Apply and restart UI” on the “Installed” tab to restart the WebUI to ensure that the extended functions take effect.

Installation: Photoshop

1. in the plugin’s GitHub page Move to “Method 1: One Click Installer”, click on the “.ccx” link;

2. Download the .ccx one-click installation file on the release page. Version 1.2.4 at the time of writing;

3. After double-clicking the downloaded .ccx file, Creative Cloud Desktop will be automatically launched, and the program will prompt the user to note that the plug-in is an unofficial plug-in, and press the “Install” button to continue;

4. Press “OK” to agree to the installation, then Windows will pop up a system prompt, press “Yes” to agree to the installation;

5. After the installation is complete, it will be displayed in Creative Cloud Desktop, and Photoshop can be opened directly.

6. Be careful when creating new filesDo not check “Working Artboard”and the size should be generated from a small size, and can be enlarged with Stable Diffusion when modifying afterwards.

7. You can find the newly installed Auto Photoshop Stable Diffusion Plugin in “Main Menu > Add-Ins”.

Open the Auto Photoshop Stable Diffusion plug-in, and you will see its interface on the right. There are 6 tabs in it, including pictures and prompt records generated in the file, Controlnet settings, Extra for AI enlarged pictures, etc. You can also choose different models and LoRA, set positive and negative prompts, choose Vincent graph, graph graph, Inpainting or Outpainting, generate multiple graphs continuously, set the generation size, number of steps, and sampler, etc., and also have appearance correction and high-resolution correction option.

In terms of use, Photoshop embedded in WebUI can make full use of the strengths of both parties. The generation of WebUI is flexible enough, and the movement of characters can be controlled through Controlnet. At the same time, Photoshop’s functions of fading, filters and layer effects can be used for creation. Not a bad combination for creators. However, because both programs consume a lot of resources, they also have considerable requirements for computer specifications.