I have created a set of poses using the openpose tool from the Controlnet system. The set consists of 25 unique poses, each with 25 different angles from top to bottom and right to left.
All the images in the set are in png format with the background removed, making it possible to use multiple images in a single scene without worrying about overlapping backgrounds. Additionally, the Controlnet system does not require a background to use the images, so they can be imported as they are.
However, some poses may be too far from the scene and require zooming in for stable diffusion to capture them properly.
Some poses in the set, such as those with extreme angles like top-down or bottom-right upward views, may be challenging for stable diffusion to comprehend without additional prompts support. If stable diffusion is unable to understand or generate these extreme poses, it is advisable to move on to other poses.
Overall, have fun experimenting with the pose set, and feel free to share your thoughts with me.
Description
FAQ
Comments (8)
We have "poses" tag for this things, please use it
Thank you for reminding me about the tag for posing. I will make sure to use it for the upcoming packs :)
How would I use this? New here sorry
You need to have a extension call Controlnet, you need to install it and use the poses in Openpose option.
I have recorded a video showing how I use these poses. and some solutions for issues you might encounter.
https://www.youtube.com/watch?v=-bsGo3bSx0M
Great set of poses, the angles are so useful. One question: any tips on ensuring read angles render as rear shots and not just front shots in different positions?
Not sure I understand what you are asking, but I have recorded a video showing how I use these poses. and some solutions for issues you might encounter.
https://www.youtube.com/watch?v=-bsGo3bSx0M
How should I use these actions? put them in that file?