https://buymeacoffee.com/Kraz3yivan
The rough logic of the mask is it rotates the image and uses three (technically six) comb detectors to roughly find dot crawl, and it uses an alternative 1x AI processed clip to gen a chroma edge mask to limit the detection to chroma edges.
The rough logic of the mask is it rotates the image and uses three (technically six) comb detectors to roughly find dot crawl, and it uses an alternative 1x AI processed clip to gen a chroma edge mask to limit the detection to chroma edges.
Some testing with Video Depth Anything here on "Q-less" S1
#DS9 - the aim is to layer various masks and AI upscales to almost emulate photoshops layers but with video 😄
Check it out here...
github.com/DepthAnythin...
Some testing with Video Depth Anything here on "Q-less" S1
#DS9 - the aim is to layer various masks and AI upscales to almost emulate photoshops layers but with video 😄
Check it out here...
github.com/DepthAnythin...