Change Poses of Existing Images! Stable Diffusion & ControlNet Tutorial

Stable Diffusion Tutorials & More

Links 👇

Try Stable Diffusion In The Cloud 👇

3 Budget Graphic Cards for Stable Diffusion👇

3 Budget Desktop PCs for Stable Diffusion👇

3 Budget Laptops for Stable Diffusion👇

Software 👇

🔔 Subscribe now with all notifications on for more videos about The Newest AI, Stable Diffusion Tricks & More!
#stablediffusion #openpose #controlnet #posing
📺 Watch the entire video for more information!
_________________________________________________________________
About Next Diffusion
🎥 Videos about AI Image Generation, AI Video Generation, Stable Diffusion & More.
🎨 Written, voiced and produced by Next Diffusion
🔔 Subscribe now for more videos about The Newest AI, Stable Diffusion Tricks & More!

Watch More from Next Diffusion

_____________________________________________________
💼 Business Inquiries and Contact
• For business inquiries, copyright matters or other inquiries please contact us
❓ Copyright Questions

Timestamps:
0:00 - Intro & Examples
0:25 - Requirements
1:08 - Txt2img Settings
2:16 - ControlNet Settings
3:33 - Example
4:00 - Subscribe!

[kicapanmanis]

So I followed the video instructions using my own image and did a few attempts and my end pose doesn't look like the openpose picture at all. Even used two pictures (one reference and one openpose), which some others commentors did not initially.

Followed the instructions for both controlnets to the latter. Downloaded and use your pose files from CivitAI for the second Controlnet. It just doesn't want to follow the pose. Tried using my original prompts from when I first generated the image before this as well as the Danbooro one, neither worked.

Used the None preprocessor for second Controlnet and the openpose model you asked to download. Reference image in first Controlnet is set to reference as well. It just doesn't want to copy the pose. Multi Controlnet is enabled in settings.

[erickromano]

why it always get black photos as result?

[Eeatch]

How the hell did you learned all of it??? Nice video, and i'll definetely be checking out your whole playlist

[VKS]

Thank you for your great tutorial.
I've got the issue that my pictures get very dark. I guess that's because of the black background of the pose.
How can I fix this?

[somyeol]

awesome tutorial, i was trying to find a good tutorial and here it is!

[mateuszlobien]

Interesting, can you mix two images with different characters then pose models in order to mix the two characters in to one image. Exemple two characters from two different pictures resulting in these two riding a bike ?

[h.e.rimaginglidiankalter]

This is great, how can I get this on my Apple system?

[lugotorix]

Thank you for the tutorial, I managed to get an image in a different pose with the method you mentioned, but when I wanted to change the pose, it continued to reproduce the previous pose exactly. Then, I tried sending the prompts from the png info again as you showed, but it did not work. Do you have any solution suggestions?

[Zielak]

@nextdiffusion of my understanding, here you are providing pose model, but could it be posible that you could change pose of existing image using nothing but descriptive prompt about the pose, possibly adding foreign objects to the image? like "sitting with arms crossed with a baseball hat"

[rohitstube]

Thanks @NextDiffusion making such nice tutorial, i had one question, while using realistic vision 5.1 checkpoint followed by two control net reference then pose, for my own image, quality of image are not consistent across face and pose, can you provide your suggestions?

[JipandCie]

awesome ! Do you know how doing the same workflow in comfyui ?

[Gillz]

Help, I am doing everything you are exactly doing. I have downloaded everything you have shared in the video. The problem is, my character keeps changing. I get the pose i am needing, but my character keeps changing. Its so frustrating.

[skroudge]

I want to make my own manga by ai but I want consistency in characters so no I can do that

[riyhype]

I'm new and have just started using SD via a sagemaker labs notebook, but I can't get dynamic poses to work. I think the issue is the sd15 model isn't one of the options. I've tried everything and nothing seems to work.

[RandomInternetDude]

What if my controlNet doesnt have window to drop in a picture :S

[made]

Can you explain why the pose is not the same at the end?

[youbetterwork-LT]

Does this work with stable diffusion Forge???