News for hackers. Stuff that matters.

A comment on PixelNN – Example-Based Image Synthesis. Parent comment

Re:PixelNN – Example-Based Image Synthesis by amelius on Monday September 25, 2017 @08:03AM (#15329673)
> It's similar to that guy who used Photoshop's swirl effect to hide his face, not knowing that the effect is deterministic, and can easily be undone.

The effect does not only need to be deterministic, but also invertible.

A low-res image has multiple "inverses" (yikes), supposedly each with an associated probability (if you would model it that way). So it would be more honest if the algorithm shows them all.

[ Parent | on HN ]
Re:PixelNN – Example-Based Image Synthesis by tentaTherapist on Monday September 25, 2017 @03:20PM (#15333208)
Showing them all seems a bit impossible because the number would blow up really quickly, wouldn't it? Maybe it could categorise them, but that could be misleading, too... I don't know.
[ Parent | on HN ]
Re:PixelNN – Example-Based Image Synthesis by pedrosorio on Monday September 25, 2017 @12:26PM (#15331746)
It's what they call an https://en.wikipedia.org/wiki/Inverse_problem
[ Parent | on HN ]
Re:PixelNN – Example-Based Image Synthesis by tentaTherapist on Monday September 25, 2017 @02:58PM (#15333003)
That is a very well-named problem.
[ Parent | on HN ]