openpose-editor VS MultiDiffusion

you can try Openpose Editor editor and make a small skeletons with a large black image and generate a pixel perfect image with the same size as the large one to give the controlnet a perspective of a small creature

You want the Openpose Editor extension.

Le message d'erreur est: GitCommandError: cmd ('git') a échoué en raison de: Code de sortie (128) CMDLINE: GIT CLONE -V - https://github.com/fkunn1326/openpose-editor.git C: \ Users \ user \ stable-diffusion-webui \ tmp \ openpose-editor stderr: 'fatal: path de destination' c: \ utilisateurs \ user \ stable-diffusion-webui \ tmp \ openpose-editor 'déjà existe et n'est pas un répertoire vide. '

I wanted to use OpenPose as well, however the preprocessor did not want to recognize the exaggerated cartoon. So I pulled her into OpenPose editor and traced the skeleton, putting her hand behind her head since the original was weirdly posed anyway. Exported the PNG, and brought it into the second CN slot, set to the openpose model with NO preprocessor. An ideal weight turned out to be 1.5. I chose to let the prompt be more important (old guess mode) on both CN inputs.

Error message is : GitCommandError: Cmd('git') failed due to: exit code(128) cmdline: git clone -v -- https://github.com/fkunn1326/openpose-editor.git C:\Users\user\stable-diffusion-webui\tmp\openpose-editor stderr: 'fatal: destination path 'C:\Users\user\stable-diffusion-webui\tmp\openpose-editor' already exists and is not an empty directory. '

The seed is only going to impact the "random" stuff. You would need to change the ControlNet input. OpenPose editor can load an image and generate a pose for ControlNet but you will need to edit it to reflection the change of perspective. There are some 3D options among the extensions as well but you would probably need to start from scratch with those. https://github.com/fkunn1326/openpose-editor

I've installed the auto1111 openpose editor from https://github.com/fkunn1326/openpose-editor.git and the control net gui from https://github.com/Mikubill/sd-webui-controlnet.git. However, upon launch the error message above is given and the openpose editor isn't there. How do I fix this issue?

!git clone https://github.com/fkunn1326/openpose-editor /workspace/stable-diffusion-webui/extensions/openpose-editor

Afterwards I put it into img2img and played around with the open-pose editor First I recreated the original pose, which went surprisingly well. Then I just had to create the poses for the hip swaying. Played around a bit with denosing strength aswell, I was mostly at 0.7-0.85. I put the poses into ControlNet one after another and this was the result. Could be made a lot smoother with more images, as I only used I think 10 different poses, but this was just a quick study for me on how well you can move a character with ControlNet. Using the correct prompts it could definetly be possible to spin smoothly spin the character around, so I might give that a shot later. But for now I'm once again off making LORAs and the occasional "normal" post. See ya

It works now after reinstalling openpose, i use this one: GitHub - fkunn1326/openpose-editor: Openpose Editor for AUTOMATIC1111's stable-diffusion-webui

For composing this approach works pretty well

https://multidiffusion.github.io/

So I've continued to experiment with how many papers I can fit into a single pipe and have them play nicely together. The images below were created by combining the panorama code from omerbt/MultiDiffusion with the ideas from albarji/mixture-of-diffusers. Also turns out nateraw/stable-diffusion-videos can be seen as a special case of a panorama (in latent space rather than prompt space).

Essentially, I fine-tuned Stable Diffusion 2.1 base (the 512x512) model on the ESA Hubble Deep Space Images & Captions dataset I collected from public Hubble images & captions. After around 33,000 training steps, I saved the model and was really impressed by the results. But I really wanted to be able to generate wallpaper-level quality space images, so I stumbled upon MultiDiffusion: a new project for generating massive panorama images using stable diffusion models. I then used hubble-diffusion-2 along with MultiDiffusion to generate each one of these amazing 2560x1536 images. Each image took a little over an hour to generate on a Google Colab T4 GPU. I used the following prompts for each of these images:

If you need generated and not upscaled 4k for some reason, try something like https://github.com/omerbt/MultiDiffusion

Project: https://multidiffusion.github.io/ Paper: https://arxiv.org/abs/2302.08113 GitHub: https://github.com/omerbt/MultiDiffusion

Quick Read: https://www.marktechpost.com/2023/02/24/meet-multidiffusion-a-unified-ai-framework-that-enables-versatile-and-controllable-image-generation-using-a-pre-trained-text-to-image-diffusion-model/ Paper: https://arxiv.org/abs/2302.08113 Github: https://github.com/omerbt/MultiDiffusion Project: https://multidiffusion.github.io/

So I haven't made many images with Stable Diffusion despite using it heavily. The reason is I've been messing with the internals of the diffusion pipe, to interfere with the diffusion process in different ways. Todays fun result is based on omerbt/MultiDiffusion for making panoramas.

If we combine Mixture of Diffusers + MultiDiffusion+ Composer+ cross-domain-compositing and probably some more I'm not thinking of.