"Optimize AI Art Creation with ControlNet and ComfyUI: Explore an Amazing Automated Workflow"
HTML-код
- Опубликовано: 28 июл 2024
- An overview of the workflow concept from 30:52
"This video delves into the diverse Controlnet options available, including Canny, Scribble, Openpose, and Depth, as well as the new SDXL Controlnets from xibsir. As always, our dynamic duo, the Ziggys, presents these features in an entertaining and engaging way, ensuring that viewers not only learn but also have fun along the way."
CivitAI V5.01:
Minor Release - Usability - Model preselection and Controlnet preview images can be switched off
civitai.com/models/555022
Openart:
Upload doesn't work. Maybe tomorrow...
0:00 Start and welcome of the audience
0:29 Presentation of the Controlnet groups
3:29 Complete overview of preprocessors
4:34 Restart in training environment
5:00 Installation of preprocessor custom nodes
6:47 Running first test without Controlnet
7:35 Preparing for Controlnet catwalk
8:50 Download Controlnet models
10:10 Restart with Controlnet models
10:22 Select canny model
11:00 Example of deleting groups from the workflow
11:52 First run with Canny
13:00 First run with Open Pose
14:50 Second attempt with other Open Pose model
15:55 Info new XinSir SDXL Controlnet models and download
17:20 Explanation of the different XinSir SDXL Controlnet models
19:05 XinSir Openpose Controlnet test
20:36 Continuation after restart with different examples
24:30 New xinSir Depth Model Various great examples
25:32 Examples combining controlnet and image to image
29:55 Ballerina in dark gothic Masterpice
30:55 Repetition of all important settings of the entire workflow 
Наука
The last section of the video should be tagged at the top of the homepage, which not only contains an introduction to all functions but also gives viewers an overall concept of your dedicated workflow.
That's a good idea, I pinned your comment. It starts at 30:52
Fantastic work Murphy. Looking forward to watching this fully after work. Like the new voice too 😊
@Dreddoom ...neue Stimme und sie drückt auch noch England die Daumen 😲
Ups, I wanted to say, new voice and she's still rooting for England too😮
😊 That's good to hear
Thank for making this. I really enjoy this detailed presentation and for sure will explore and tinker with the workflow! Inspiration is one thing but creativity is another thing :)
@demoscenes Thank you for your comment and have fun tinkering ;-)
Does the xinsir's union model work as well as the other respective ones(xinsir's)? If it does, that means we need to switch to specific xinsir models; for example, select a xinsir's depth model in the depth section of the workflow.
That's a very good question and I can't answer it seriously. I don't use Controlnet that much because it slows down image generation on my laptop. I probably wouldn't switch again, especially if I get what I want to achieve. Small info: There is another newer all in one model from Xinsir; diffusion_pytorch_model_promax.safetensors which can also be loaded via the manager.
@@murphylanga I agree with you because the image2image feature has already done a decent job, which sticks to the image by adjusting the denoise value.
Despite being time-consuming, a myriad of exquisite models in Controlnet is still very exploratory.
Thank you for your reminder; I will check the new model later.
Really enjoy your work Murphy! Learning a lot just by tinkering with all the gadgets and gizmos. Quick question - I noticed nodes are not pinned, is there a quick way to pin them all at once? I end up moving them around by accident. Thank you!
@bobbobster8619 Oh sorry, I forgot to pin the nodes. In the information window. You can find it if you press 0. Is this described:
Nodes are not pinned
Use github.com/blib-la/blibla-comfyui-extensions to pin or frezze the nodes and groups! You can also load the nodes via the manager.
@@murphylanga Thank you !
I just wanted to Thank YOU! You inspired me to start my own youtube channel. You also helped teach me how to make the beautiful things I love staring at all day! What a time to be alive!
Wow, that's amazing to hear! I'm absolutely thrilled that my tutorials have inspired you to start your own RUclips channel and create beautiful AI-generated art. It's truly a joy to see people unlocking their creative potential and embracing the power of technology in art. I'm so grateful for your kind words and wish you all the best on your artistic journey. Keep exploring, learning, and creating-the possibilities are endless! 🎨
Hi ! I wanted to know if you have any idea how zia fit on I.ns.sta is made ? It's a specillized ai agency behind and i don't know what kind of workflow they use to get this level of quality, consitency.
It seems that the base image is an exsiting one, but then they maybe use a 3d pose character + openpose + lora for body + lora for face, but something is off. I you have any idea on how to do that it would be really interesting to know, they are kinda the best in their field.
I don't know Zia Fit and I don't use Insta either. I had a quick look to see what it could be. Looks like they use the Face Raactor Node or something like that. There is also Roop under Automatic1111. Both are not released for commercial use!
Maybe there is something similar, or better, for which you have to pay license costs. The faceswap model used in the Reactor Node is deliberately "throttled" to prevent abuse. But even if you use it, read the disclaimers and license terms carefully!