Using AI to create multiple images of the same person from one image (for comics)?

hochkampByhochkamp

I want to create images for a comic using artificial intelligence. I used Canva to create a watercolor-style image of a cute police cat that I really like. Now I want to somehow get an AI to create 10 more images from this one image, each with the cat doing different things, like riding a bike or shopping, etc. But I want the cat to always be the same. Does anyone have any ideas on the best way to do this? P.S. The attached image is an example so you can get an idea of ​​what kind of image I mean.

(2 votes)
Loading...

Similar Posts

Share local website on the Internet?

ByKlaus5624

Hello everyone, I have created a local website that I would like to publish on the Internet. With Xampp. now the following error message appears klaus1.ipv64.net refused the connection. Whether ipv64 or spdyn, the same errors always occur. I created the site with Joomla. My old computer crashed, now I have a Win 11 computer….

change the Y value to jump?

ByITTimo

Script: public float speed = 5f; private Rigidbody2D Player; public float jumpForce = 5f; void Start() { Player = GetComponent<Rigidbody2D>(); } void Update() { float h = Input.GetAxis("Horizontal"); if(Input.GetButtonDown("Jump")) { Player.AddForce(Vector2.up * jumpForce, ForceMode2D.Impulse); } Vector2 vec = new(h, 0); Player.velocity = vec * speed; } First of all, I'm a beginner programmer and wanted…

Use pixelmatch in browser?

Byjumbo125

If I use the link on the github page: Code: <script src=" https://bundle.run/pixelmatch "></script> no file is found, the path does not exist if I download and integrate the latest release Code: <script src="myserver/js/pixelmatch/bin/pixelmatch"></script> I receive the following error message Code: const PNG = require('pngjs').PNG; ReferenceError: require is not defined

Subscribe
Notify of
2 Answers
Oldest
Newest Most Voted
Inline Feedbacks
View all comments
M0din06
M0din06
1 year ago

This is relatively difficult. Of course, it is possible to create the same or at least similar person with a ‘text-to-image’ AI. However, this with only one existing image becomes quite difficult.

Such AI models are usually fed with hundreds or even thousands of images of the same person until the artificial intelligent knows each individual detail of the person and can thus create very similar photos.

In a picture like that of your cat there are not too many details, but it would still be recommended to have at least 50 different pictures for “feeding” the AI.

In addition, you need (possibly) the know-how to train such a AI. While there are now some videos on YouTube (etc.) and even websites that are specialized in it, most of them cost money.

I hope I could help you with your question a little bit.

4
helpedbyanerd
helpedbyanerd
1 year ago

You could download a suitable model from civitai.com and create the images with Automatic1111. The Adetailer extension is good for very similar faces. “Perfect” it becomes rare.

1