Disclaimer:
The installation package used in this case is provided free of charge without any profit purpose.
Hello everyone, I am rain or shine.
In order to let everyone understand the function of generating graphs more intuitively, understand what the function of generating graphs does, and what can it do? Today we continue to introduce the practical case of Tushengtu-the production of beauty dressing.
For new friends, if you want to learn more about Stable Diffusion, please check out previous articles:
AI Painting Stable Diffusion Research (1) sd Integration Package v4.2 Version Installation Instructions
AI Painting Stable Diffusion Research (2) sd Model ControlNet1.1 Introduction and Installation
AI Painting Stable Diffusion Research (3) sd Model Types Introduction and Detailed Installation and Use
AI Painting Stable Diffusion Research (4) Detailed Explanation of sd Wensheng Diagram Function (Part 1)
AI Painting Stable Diffusion Research (5) Detailed Explanation of sd Wensheng Diagram Function (Part 2)
AI Painting Stable Diffusion Research (6) sd Prompt Word Plug-in
AI Painting Stable Diffusion Research (7) ) One article to understand the working principle of Stable Diffusion AI painting Stable Diffusion
Research (8) Detailed explanation of sd sampling
method
Detailed explanation of the function of generating pictures - making beautiful QR codes
The old rules, here are still explained in the way of case interspersed with knowledge points.
Dear friends, you can follow my steps while doing practical operations and learning theoretical knowledge at the same time, which can improve learning efficiency.
Beauty Dress Up Production Case
What should we do if we want to achieve the same effect as the picture above?
The first step is to prepare the beauty picture information and model
Here we take Hanfu style as an example to illustrate.
(1) We can go to Station C and search for "Hanfu"
as the picture shows:
According to the parameters of the picture, we found that the main model uses Chilloutmix-Ni-pruned-fp32-fix, and of course the Lora model hanfu2.0, if you do not have a friend installed locally, please download and install it first.
The model can be downloaded directly at station c, so I won’t detail it here.
If the download is slow, here are the models used in the case .
(2), install the model
After the download is complete, put the Lora model file into the models\Lora directory
\sd-webui-aki-v4.2\models\Lora
Put the Chilloutmix-Ni-pruned-fp32-fix model file in the models\Stable-diffusion directory
\sd-webui-aki-v4.2\models\Stable-diffusion
After the installation is complete, remember to restart the console and webui to make the model take effect.
(3), copy and set the picture parameters
-
Copy the forward prompt words and reverse prompt words on the right side respectively, and paste them into the prompt word input box of the webui map
-
Main model selection Chilloutmix-Ni-pruned-fp32-fix
-
CFG Scale: 8
-
Number of iteration steps: 40
-
Sampling method selection dpm++ 2s a karras
-
Random seed setting: 2589415615
The parameters of sd webui wensheng graph are set as follows:
The second step is to use the Wensheng map function to generate a beautiful picture and send it to redraw
After the above parameter settings are OK, click the "Generate" button directly to send the generated picture to redraw.
The generated image is as follows:
Click the "Send to redraw" button, and at this time, it will automatically switch to the redraw interface of the graph generation graph.
The third step is to select the brush icon on the right and paint the area that needs to be changed
ctrl+mouse wheel to adjust the thickness of the brush.
The fourth step is to modify the prompt words in the forward direction, and modify the prompt words related to clothes
The positive prompt word before modification: The clothes are Mingfu, Mingzhuang.
(8k, best quality, masterpiece:1), (realistic, photo-realistic:1),cute, dating,(smile:0.9), (closed mouth),small breasts, beautiful detailed eyes,<lora:hanfu2.0:1>, jin, (collar button, ming, Ming Dynasty, Ming clothing:1.4)
Here we change clothes into Tang suits, Tang suits.
Modified prompt words:
(8k, best quality, masterpiece:1), (realistic, photo-realistic:1),cute, dating,(smile:0.9), (closed mouth),small breasts, beautiful detailed eyes,<lora:hanfu2.0:1>, jin, (Tang suit, Tang Dynasty, Tang suit: 1.4)
The fifth step is to keep the gestures and shapes of the characters unchanged, which need to be controlled through ControlNet
-
Enable ControlNet
-
Control type selection: OpenPose
-
Preprocessor: openpose_full
-
Model: control_v11p_sd15_openpose
The sixth step is to generate the viewing effect
Let's change the style of a Hanfu again, only need to modify the prompt words related to the clothes:
(8k, best quality, masterpiece:1), (realistic, photo-realistic:1),cute, dating,(smile:0.9), (closed mouth),small breasts, beautiful detailed eyes,<lora:hanfu2.0:1>, jin, (Hanfu, Han Dynasty, Hanfu: 1.4)
Contrast effect:
Are you satisfied with the effect of this beauty's dress-up?
Well, this is the end of our third function case of graph generation.