Photo to Cartoon Stable Diffusion: Easy Step-by-Step Guide - Forge UI

Video Statistics and Information

Video
Captions Word Cloud
Reddit Comments
Captions
today I want to show you how to convert photos into 3D cartoon characters using stable diffusion this way you'll have some nice cartoon images for your social media maybe you're a designer seeking cartoon inspiration or in need of a cartoon mascot for your site or marketing materials you'll need to download a model called wild cart animation version 2 from the civit AI website click on the download button and save it in your Forge folder look for the web UI folder then navigate to models and finally to stable diffusion save the downloaded file in that folder as you can see I've already downloaded it after it's downloaded it may take a few minutes since it's quite big you can find it in the stable diffusion checkpoint drop- down menu if it doesn't appear hit refresh and then you can select that model this model is proficient at generating cartoon images for example if I try to generate a cartoon cat I get something like this click on the imageo image Tab and then upload your own image in my case it's a photo of a princess add a prompt that describes the image mentioning the objects you want to appear such as the crown and the veil and then to ensure it becomes a cartoon add the text a professional 3D design of a cartoon followed by the subject which can be a woman car building or anything else after that add extra Pixar art style to achieve a nice clean 3D cartoon look for the negative you can find an example on the model page copy that text without the quotes and paste it into the negative prompt area then scroll down to the width and height settings and ensure they have a similar ratio to your uploaded image next focus on the Deno strength start with 0.35 if you want the cartoon image to closely resemble your original photo and increase it up to 0.6 for a more distinct result click on generate and and you'll receive a cartoon image very similar to your original if you increase the denoy strength to 0.5 you'll notice it has more power and creates a more different version um further increasing it will will result in even bigger variations so adjust the slider until you achieve a result you like keep in mind that some images are easier to convert into a cartoon version than others so experiment with different prompts Deno strengths and images I'll remove the current image and upload a photo of a dog instead for the prompt I'll adjust the first part to say a dog and replace woman with dog after hitting generate we'll get a very similar cartoon version of the dog when dealing with animals I like to increase the denoy strength a bit more around 0.45 to achieve a cool cartoon look with big eyes and a nice 3D appearance you can generate the image a few times until you find one you like now let's convert an object like a car this time I'll change the prompt to a white car despite the misspelling it will still work since I want the objects to be even more exaggerated than the animals I'll use a Deno strength of 0.6 resulting in a really cool 3D cartoon car um I'll try the next photo with a girl and remember to adapt The Prompt accordingly to avoid mutations since people typically need a lower denoy strength compared to animals and objects I'll adjust the strength for the girl to achieve a similar look you can even drag the results and use that as a starting point to get a more cartoonish version while still resembling the original one nice feature is that you can adjust the prompt to include more details for example I can make the girl Angry by adding angry to The Prompt or I can make her cry let's increase the denoise to get an exaggerated version of that cartoon you can have fun playing with emotions adding different colors objects and so on here's a quick example of a lion and how I obtained a cartoon version of it you can also try with buildings however to achieve a truly cartoonish version uh you'll need to increase the denoy strength but this may make it less similar to the original so it's a bit of a compromise if it closely resembles the building it won't have uh the exaggerated features characteristic of a cartoon you can also experiment with using control net uh some of you have asked for control net videos but I'm still encountering errors when the image size is not divisible by 64 therefore I'm waiting for a fix before creating a tutorial for control net it also works with plants and trees so you can get really creative with images around you I'll quickly try another one for a man with a beard notice how I play around with prompts and denoise strength experimentation is key to getting what you want try different seeds change the prompt add different colors and lights usually cartoons are more colorful and have exaggerated features so make it fun speaking of fun I took a photo of a measuring tool with my fingers to create a body-like shape um I use this prompt for a cartoon robot with a denoy strength of 0.6 to make it more creative as you can see I can generate a few times to get different toy robot ideas based on that simple image I mean how cool is that I mean only your imagination is your limit I took a brush from the kitchen and turned it into a monster this tool could be really helpful for concept artists to come up with new original ideas based on a random image if you want more control over how things are generated you can do it like I do in Photoshop you can quickly paint the shapes and colors your cartoon images should have for example here I used a hard transfer brush to quickly sketch the shape of a monster head it took less than a minute to create a rough sketch squint your eyes and when you can see the basic shapes you can copy that image and paste into stable diffusion here you just uh add a prompt that describes what you want to create uh and with a denoy strength of 0.5 in my case you can generate a few versions to see what you can come up with I just noticed the initial sketch was too small on the canvas so I enlarged it in Photoshop to fit better then I pasted it back into stable uh stable diffusion and generated it again to get a nice quality cart Tune monster yeah it's a unique design in Just 2 minutes I'll use the same technique to create the YouTube thumbnail I hope I've given you some ideas on how you can approach cartoon creation if you found this video helpful leave a like thank you for watching and for your support have an awesome week
Info
Channel: pixaroma
Views: 3,823
Rating: undefined out of 5
Keywords: photo to cartoon ai free, photo to cartoon ai, image to cartoon ai, image to cartoon stablediffusion, image to cartoon character, photo to cartoon ai tutorial, forge ui stable diffusion, forge ui tutorial, forge ui img2img, stable diffusion tutorial, stable diffusion, ai art, stable diffusion ai, 3d cartoon character ai, stable diffusion ai art, stable diffusion art, pixaroma, stable diffusion web ui forge, stable diffusion img2img, stable diffusion webui forge, cartoon
Id: 0ReO_RInX1E
Channel Id: undefined
Length: 7min 32sec (452 seconds)
Published: Mon Apr 08 2024
Related Videos
Note
Please note that this website is currently a work in progress! Lots of interesting data and statistics to come.