Ai feedback/ideas

[Nikita29167753hl2z]

Hi there,

I absolutely love generative ai in my workflow. Thank you for developing it! I'm working as a concept character/creature artist in film industry. The ability to select my character and generate a background that works with the camera/perspective is insanely powerful. It's the feature I will absolutely use in the future. It's something I've always wished I had. There are a few things I would love to have this feature to have:

1) When background is generated, it also tries to fill in some areas of my character. This results in changing the shape of the head, weird stuff with fingers. I think ideally it would create purely a background in this case. This way if needed - I can flip it easier. I can blur  it and not affect the edges of my character.

2) Would be great if I like a background - save it as a "working version" or something. So i can build upon it. For example if I have a few shots of the same character - I can use the same background but rotate it and adjust slightly with different character shots. Or say I like everything in bg, but want to add a light source or an element to the right side of the screen. So i could just build upon the image I already like.

Thank you for reading through,

Nikita

[Nancy OShea]

I'm working as a concept character/creature artist in film industry. The ability to select my character and generate a background that works with the camera/perspective is insanely powerful.

By @Nikita29167753hl2z

===========

Yes.  But beta products are not for commercial use. 

"During the Firefly beta, any Firefly-generated assets cannot be used for commercial purposes."

More importantly, adding generative AI content to your work significantly effects its protection under current copyright laws. 😖

https://www.artnews.com/art-news/news/ai-generator-art-text-us-copyright-policy-1234661683/

[Nikita29167753hl2z]

Yeah I get it. So far just tried out with a few of my old personal projects. Once the ai stuff gets ironed out with all the laws then of course it will be available for commercial use. I'm absolutely against AI generating a finished art piece. I'm an artist myself and I've been working really hard for the past 13 years to be where I am right now. However if we can use AI as a supporting tool in our art projects - I'm all for it. I personally can see an option where say those background images were generated from a pool of stock photos access to which I'm paying - that's fine by me and will be fine by the artists who contribute their work to the stock photos and get paid out for having their work used in generation of such content. 

[Matthew249224202mcs]

Something that I think would be super useful is a way of saying "Don't change this" 

As Nikkita mentioned, I surround my subject with a bounding box, inverse it to get the background and then generate... 
In almost every circumstance, it will take an object the model is holding and expand on that in the results.  The only way I've found to get around this is to slice up the picture and then reassemble it + clean up, which takes much longer.

If I could paint a certain part of the image as  "don't consider or change this" I think it would make things a lot easier.