toyxyzさんのイラストまとめ

@bilawalsidhu Normal pass doesn't seem to work as well as depth. The top is a Blender rendered normal, and the bottom is an automatically generated normal.

#10_ Overall Workflow

#07_ ABG Remover Script makes it easy to remove the background. We need to tweak the generated mask a bit to make it cleaner. https://t.co/kOpg9djsoF

#05_ When using the Img2Img tab, it is possible to adjust the lighting of the background by adding a color image to the img2img. Even when the denoising strength is high (1.0), it is affected by the brightness and tone of the input color image. 1 : txt2img / 2,3,4 : img2img

#03_ The reason for using the rendered depth rather than the automatically generated depth is that the sharper the boundary of the depth, the better the detail and sharpness of the image. Top: Rendered depth / Bottom: Auto-generated depth

ControlNet Depth&pose test #stablediffusion #AIイラスト #pose2image #depth2image

ControlNet Depth2image test #stablediffusion #AIイラスト #depth2image

#07_ And render the image sequence >BatchImg2Img. To run BatchImg2Img using ControlNet, you must enable "Do not append detectmap to output" in settings.

#06_ And of course, these openpose-like bones can also be retargeted to Mocap. I used Autorig-pro. https://t.co/d5y5izbLn3