A node that applies ML-based technologies to remove backgrounds from photoes.
And here is a question regarding a possible way to achieve that.
Assuming that in Developing a Library Module API section is stated that
A library module can be implemented in C, C++, Objective-C, or another language, as long as it exports C symbols.
is a reasonable path the idea to port a swift code like:
https://gist.github.com/vlondon/491c2e7829d60e835d53a1f6810a34ed building as bc and then creating a node based on that? Or is it sci-fi?
Yeah would love to see more of this kind of stuff too :)
Your ML model seem to be about general background removal, I asked here on (Questions about ML models & techniques for feature requests) for people segmentation too.
We’ve opened this up for community voting.
1 Like
is a reasonable path the idea to port a swift code like: https://gist.github.com/vlondon/491c2e7829d60e835d53a1f6810a34ed building as bc and then creating a node based on that?
Possibly you could create an Objective-C wrapper that calls the Swift code. Or, in this case, the Swift code is short enough that you could perhaps rewrite it in Objective-C.
Another option is VNGeneratePersonSegmentationRequest, new in macOS 12.
Here you mentioned https://github.com/andrewssobral/bgslibrary. Since that one depends on OpenCV, which is a large and complex library, it sounds potentially more difficult to port to Vuo than the above options (as you may have already noticed).
1 Like