Create Consistent Characters from ONE Image with OpenArt

OpenArt released their Characters feature that lets you create a character from a description, single image, or 4+ images. The folks at OpenArt have been working hard to develop the Characters feature – not only tweaking things in the backend to just make it work better – but adding a bunch of features for us to use for creating and using our characters.

Even though you can create a character from one image in OpenArt, and then use that character to generate other images (and then convert those to videos, if you want), the option to create a character from 4+ images is there for a reason… providing more high-quality images of your character going into the character creation tool usually means more consistent, better quality results when generating images with your character.

In this video, I’ll show you how I took one single image of a character and turned it into 6 images of that character, with different backgrounds and clothing, so I could create a character from 6 images instead of 1 – giving those AI robots more to work with and more details about my character.

I’ll explain how I take a single image of my character and:
-Generate a video that has the character turn and change facial expression
-Extract frames from that video (individual images at different intervals in the video depicting my character from slightly different angles, with varied facial expressions)
Then, with those individual frames (images)
-Use Inpaint in the OpenArt Editor to change the clothing
-Use the Blend Board in the OpenArt Editor to change the background
End up with 6 images of my character in different clothing and a variety of backgrounds, which can then be used as the training images to create a Character in OpenArt.