Generate Cartoon Character

Function

You can generate a Disney cartoon style character using your own face.
It can also be conducted via numerous AI applications on phones, however, those applications usually require a monthly subscription fee.
The advantage of using Stable Diffusion is to save the subscription cost.

Procedures

​​1. Download and Use the mobile APP Voila and select “2D cartoon” to convert a real life portrait into a cartoon image.
Sometimes it might take a few attempts to generate a favoruable result for the later processes.

​2. Once you have obtained an desireable image, Open Stable Diffusion and Select the checkpoint “helloip3dIP_helloip3dV10f” on the top-left corner.
If you haven’t installed the checkpoint, please download the checkpoint file on Civitai and put it into your stable diffusion folder.

​​3. Input the following settings:

4. For Prompts, you may use the following to start with:

Postive prompt ( you may includes the following keywords):

(masterpiece), (best quality)

​Negative prompt:

bad-hands-5, ng_deepnegative_v1_75t,(extra fingers, deformed hands, polydactyl:1.4), (low quality, worst quality:1.4), (monochrome:1.1),(greyscale), watermark, text, blurry, jpeg artifacts, cropped, normal quality, signature, username, artist name, cartoon, canvas frame, lowres, (disfigured), (bad art),(deformed), (extra limbs), (b&w), weird colors, (duplicate),(morbid), (mutilated), mutated hands, (poorly drawn hands), (poorly drawn face), (mutation), (ugly), (bad proportions), cloned face, out of frame, gross proportions, (malformed limbs), (missing arms), (missing legs), (extra arms), (extra legs), fused fingers, (long neck), lowres, (grayscale), (skin spots), acnes, skin blemishes, (age spot), badhandv4

5. To rescale the image, go to the “Extras” tab of Stable diffusion

​6. ​​​Upload the result image and change the resize number to the scale-up the size you want. For example, if your original image is 512*512 and you resize by 2, the output image will be 1024*1024. You may also follow the settings below for upscaling image:

7. If you want to generate a body for the character, use “txt2img” function.

​8. Input the following prompts to generate a cartoon body for the character

​Positive prompts:

((1boy)), ((casual deep blue T-shirt:1.3)), (best quality, masterpiece), ((full body:1.3)), Diverse body types, white background, short hair

Negative prompt:

bad-hands-5, ng_deepnegative_v1_75t, (extra fingers, deformed hands, polydactyl:1.4), (low quality, worst quality:1.4), (monochrome:1.1), (greyscale), watermark, text, blurry, jpeg artifacts, cropped, normal quality, signature, username, artist name, cartoon, canvas frame, lowres, (disfigured), (bad art), (deformed), (extra limbs), (b&w), weird colors, (duplicate), (morbid), (mutilated), mutated hands, (poorly drawn hands), (poorly drawn face), (mutation), ugly

9. Once the head and the body is generated via Stable Diffusion, you might combine the 2 parts using any editing software that you like such as Canva, Photoshop.