Hey everybody. So, I’ve been working on some new software in my spare time. It’s been in development for a couple of years now. As somebody who has spent many years trying to build a game, only to burn out over and over I’ve changed my strategy and approach.
I started working on an image to model conversion tool. Where you simply input an image of a subject and an AI will attempt to model it in detail for you. For parts of the image that are not visible, like the backside of a car for example, it will creatively fill in those gaps while trying to maintain visual coherency. It does not texture models though. It simply provides an .obj file for use in your preferred 3D software.
The idea is to give an artist either a good starting point for their work, or if they are short on time, a finished model. Another benefit to this approach is that you can apply your creativity to a 2D space, using methods like photobashing and painting to see how the AI interprets your work. It could make for an interesting workflow. However, the motivation behind its development is mainly to speed up asset creation. Another cool feature will be batch processing. Where you fill up a directory with images and allow the AI to process each one. Modelling many objects with the touch of a button!
A lot of us have a vision or story we want to tell. Maybe a world we want to build. However, even for a skilled artist, it takes many hours of work to flesh out that world with characters and props. If we can reduce asset creation time significantly than I think creating your own world becomes more feasible.
So with all of that said, I don’t have any videos or images to share at this moment. I expect to have some in a few weeks as I work out some of the final issues with my code. I was pretty excited
about it though, and just wanted to share it with the community.
Ultimately I just want to make game development easier and less expensive for everybody.