Includes png and json files. Choose below.

Download the zip file. Inside are 20 .png and .json files used for poses. Requires the ControlNet extension for Automatic1111. If you don't know what those 2 things are, check out some youtube tutorials on setting them up.

A problem I've had when making preview images of my Nobodys, getting just the right zoom and position for the portraits. So I've saved me, and you some time and meticulously prepared 20 different sexy and/or practical portrait poses that the Stable Diffusion models can handle pretty well. Some easier than others. Mostly likely due to the fact that not many images of this type were trained into the model. So some models will have better results than others, and some poses are more challenging. The sample images here were generated using the legendary A to Zovya RPG Artist's Tools model.

How to use:

Download the zip file. Inside are the poses that you will instruct controlnet to use instead of generating one from an existing image. Craft your prompt as you usually would. Let the model know with prompts, this is a (sitting / laying / eyes closed) to help it along. In the controlnet settings, choose "none" for the preprocessor and "openpose" for the model. Adjust weight depending on how difficult the model wants to be. The pose images provided are 1024x1536 aspect ratio of 2:3. You can crop or add to the existing images with your favorite program to customize your composition. The .json files can be loaded with the OpenPose Editor extension and modified to your tastes.

Do you have requests? I've been putting in many more hours lately with this. That's my problem, not yours. But if you'd like to tip me, buy me a beer. Beer encourages me to ignore work and make AI models instead. Tip and make a request. I'll give it a shot if I can. Here at Ko-Fi

Search AI Models & Addons

very few results