Readit NewsDeleted Comment
The actual title of the article used the word "project" not "fabric"
Any reason to use "fabric" here?
As for RCD demosaicing, that's my next step. The color fringing is due to the naive linear interpolation for the red and blue channels. But, with the RCD strategy, if we consider that the green channel has full coverage of the image, we could use it as a guide to make interpolation better.
Consider a color histogram, then the logo (showing color oscillations) would have a wider spread and lower peaked histogram versus a correctly mapped (just the few colors plus or minus some noise) which would show a very thin but strong peak in colorspace. A a high-variance color occupation has higher entropy compared to a low-variance strongly centered peak (or multipeak) distribution.
So it seems colorspace entropy could be a strong term in a loss function for optimization (using RMAD).
Deleted Comment
1. In any case you view it, it's not trivial, which was the statement in the note. If it were, the effort to publish just for n=4 would be silly, because it would take equal effort to just publish for general case. That he withheld the proof just to mess with people is highly unlikely.
2. I definitely do not make private notes in my books just so that maybe somebody later on would pick up that book and wonder whether I had indeed discovered the secrets of the universe. I definitely do not write "challenges to the community" there.
https://en.wikipedia.org/wiki/Paradox_of_radiation_of_charge...
I believe that which is often referred to as the stagnation of physics is in a large part due to this instant-rejection in the modern physics community. There's plenty of "single point mutation" theories (think hypothesize particles with negative masses, hypothesize underlying elements below the standard model so that reactions once again obey the chemical conservation numbers,...) which individually are easy to lampoon, and are henceforth ignored (i.e. for negative masses simulations show they can pair up and accelerate indefinitely, or for a beyond-the-standard-model atomistic theory one can easily refer to the spectrum of hydrogen or positronium, and highlight that a single photon can excite it to a higher state, and then emit 2 lower energy photons).
What if our current interpretations form a very successful local optimum? I.e. suppose we can provably rule out each crooked idea if its the only modification in a theory, then we might be collectively conclude to rule them out in general, as they fail so embarassingly, but perhaps simultaneous consideration of 2 crooked ideas can make the inconsistencies disappear.
Imagine voting as a group of physicists on the most interesting crooked ideas, gathering the top 10, and then exhaustively going through the 2^10=1024 combinations, where bit K decides if crooked idea K is "enabled" a specific one of the 1024 candidates.
That sounds awesome, but it also sounds like it's conflating two things: (1) the physically impossible perpetual motion of popular understanding, e.g. machine that operates at 100% energy efficiency in perpetuity from an initial one-time energy input and (2) a machine with automatic passive energy draw from ambient sources, but with the usual inefficiencies familiar to physics and engineering.
Sounds like Drebbel did (2). Which, don't get me wrong, absolutely rocks. But I certainly wouldn't want to use (2) to advertise a moral that even laws of thermodynamics were just yet another fiction from untrustworthy institutions, which seems like the upshot you were landing on.
Suppose Fermat solved the proof by using this natural fault line -its just how this cookie crumbles- solved the n=4 case, and then smashed his head a thousand times against the problem and finally found the prime n proof.
He challenges the community, and since they don't take up the challenge, "encourages" them in a manner that may be described as trollish, by showing how to do the n=4 case. (knowing full well the prime power case proof looks totally different)
As for RCD demosaicing, that's my next step. The color fringing is due to the naive linear interpolation for the red and blue channels. But, with the RCD strategy, if we consider that the green channel has full coverage of the image, we could use it as a guide to make interpolation better.