h a l f b a k e r yWhere life imitates science.
add, search, annotate, link, view, overview, recent, by name, random
news, help, about, links, report a problem
browse anonymously,
or get an account
and write.
register,
|
|
|
I think most of us are now familiar with so-called "generative AI", where a deep neural network is used to generate text, music or images, or other data.
Considering images specifically, there are many ways this technology can be used:
text prompt -> image
image 'outpainting' (making up
data to expand the image)
image 'inpainting' (filling in areas of a partial image with made up stuff, either prompted or based on the surroundings)
AI-assisted image up-scaling
background generation
background removal
Quite often an image is close to, but not quite what is desired. For example perhaps the eyes or mouth on a face aren't the size desired, or similar.
I propose an AI-enhancement to the common graphics tool which allows an image to be distorted, either by manipulating a grid, or dragging areas in the direction they should 'spread' (called 'Warp' and 'Liquify' respectively in Krita).
This could improve the quality of the result beyond what is currently achievable.
[link]
|
|
While we're at it, can we add a "Sensible Fingers" checkbox as well? |
|
|
I've had the exact same experience which is kind of surprising. |
|
|
This has to be enacted and I assume it will. Chat GPT currently won't just take an exact image and have you be able to change just one aspect, it just starts from scratch, often wrecking the whole thing. |
|
| |