Hey everybody~!
Life lately, and the explosion of advancement in my field, have meant I don't have much time for hobbies like this. However, I noticed nobody here has played around with ControlNet yet, so I figured I'd bring it to y'all's attention.
ControlNet gives greater, well, control over your image generations, by preprocessing an image and using that data as a guide. First, I took a classic image from this subforum (IMG_1). Then I used outpainting to make the picture square (IMG_2). That was the image I sent to ControlNet's depth feature, getting the output seen in the B&W IMG_3. The AI uses this as a base, and then you can prompt on top of that base. Prompting "A cute preteen pregnant little girl wearing a business suit" gets you IMG_4, for example. "A manga cover of a cute pregnant little girl wearing a sailor uniform, my hero academia, sailor moon" generates IMG_5. "A HD professional maternity photograph of a cute pregnant little 13 year old girl wearing a bikini" creates IMG_6. "A HD professional maternity photograph of a attractive pregnant little 13 year old girl in her bedroom, wearing a lingerie nightgown" generates IMG_7.
As you can see, you can take your favorite photos and make an infinite personalized gallery of variations!
The Automatic1111 WebUI, for using Stable Diffusion at home, is available here.
The ControlNet extension for A1111 is available to download in the extensions tab of the UI or here.
ControlNet's Github page is here.
Happy Prompting!
Life lately, and the explosion of advancement in my field, have meant I don't have much time for hobbies like this. However, I noticed nobody here has played around with ControlNet yet, so I figured I'd bring it to y'all's attention.
ControlNet gives greater, well, control over your image generations, by preprocessing an image and using that data as a guide. First, I took a classic image from this subforum (IMG_1). Then I used outpainting to make the picture square (IMG_2). That was the image I sent to ControlNet's depth feature, getting the output seen in the B&W IMG_3. The AI uses this as a base, and then you can prompt on top of that base. Prompting "A cute preteen pregnant little girl wearing a business suit" gets you IMG_4, for example. "A manga cover of a cute pregnant little girl wearing a sailor uniform, my hero academia, sailor moon" generates IMG_5. "A HD professional maternity photograph of a cute pregnant little 13 year old girl wearing a bikini" creates IMG_6. "A HD professional maternity photograph of a attractive pregnant little 13 year old girl in her bedroom, wearing a lingerie nightgown" generates IMG_7.
As you can see, you can take your favorite photos and make an infinite personalized gallery of variations!
The Automatic1111 WebUI, for using Stable Diffusion at home, is available here.
The ControlNet extension for A1111 is available to download in the extensions tab of the UI or here.
ControlNet's Github page is here.
Happy Prompting!