[ImageJ-devel] halide: Decoupling Algorithms from Schedules for Easy Optimization of Image Processing Pipelines

[Curtis Rueden]

Hi Volker,

Halide looks quite cool, and is a very similar idea to other recent
libraries, including the OPS package of ImgLib, and ImgLib in general. In
the future, perhaps we could create a Halide bridge, making it easier to
invoke such algorithms on data within ImageJ. But of course it must wait
until someone is using it and/or needs it.

Regards,
Curtis


On Tue, Aug 14, 2012 at 2:22 AM, Volker Baecker
<volker.baecker at mri.cnrs.fr>wrote:

> Hello,
> I came across this work which I think could be interesting for some of you.
>
> http://people.csail.mit.edu/jrk/halide12/
>
> A citation from the side:
> > We propose a representation for feed-forward imaging pipelines that
> > separates the algorithm from its schedule, enabling high-performance
> > without sacrificing code clarity. This decoupling simplifies the
> > algorithm specification: images and intermediate buffers become
> > functions over an infinite integer domain, with no explicit storage or
> > boundary conditions. Imaging pipelines are compositions of functions.
> > Programmers separately specify scheduling strategies for the various
> > functions composing the algorithm, which allows them to efficiently
> > explore different optimizations without changing the algorithmic code
>
> Volker
>
> _______________________________________________
> ImageJ-devel mailing list
> ImageJ-devel at imagej.net
> http://imagej.net/mailman/listinfo/imagej-devel
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://imagej.net/pipermail/imagej-devel/attachments/20120816/91015e4a/attachment.html>