Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Support approximate methods for nearest neighbour color search #55

Open
adrhill opened this issue Nov 28, 2021 · 2 comments
Open

Support approximate methods for nearest neighbour color search #55

adrhill opened this issue Nov 28, 2021 · 2 comments
Labels
enhancement New feature or request

Comments

@adrhill
Copy link
Collaborator

adrhill commented Nov 28, 2021

No description provided.

@adrhill adrhill added the enhancement New feature or request label Nov 28, 2021
@adrhill adrhill changed the title Add k-d trees for nearest neighbour color search Support approximate methods for nearest neighbour color search Dec 4, 2021
@adrhill
Copy link
Collaborator Author

adrhill commented Jan 5, 2022

@johnnychen94 I've thought about this and feel like it would bloat the package too much to include a dependency on other packages like RegionTrees.jl. However we could define an interface à la "AbstractClosestColorSelector" to allow users to implement their own approximations. colordither could then have a colorselector kwarg that defaults to the current implementation.

@johnnychen94
Copy link
Member

johnnychen94 commented Jan 5, 2022

I prefer to hold this issue infinitely until people request it, or until we come up with a satisfying solution. My personal experience with nearest neighbor search is that it isn't simple at all when extreme performance is required. For instance, it took me more than a year to figure out a 5-15x solution to MATLAB on WNNM denoising, where nearest neighbour search, aka, block matching, takes about 1/2 of the computation. Yet I'm not satisfied at all so it's still experimental...

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
enhancement New feature or request
Projects
None yet
Development

No branches or pull requests

2 participants