Timeline for Is it feasible to generate every possible RGB image?
Current License: CC BY-SA 3.0
6 events
| when toggle format | what | by | license | comment | |
|---|---|---|---|---|---|
| Jan 8, 2016 at 14:29 | comment | added | zoplonix | can you prove you answer mathematically or is it just hearsay or an educated guess? | |
| Jan 5, 2016 at 19:58 | comment | added | D.W.♦ | @user1886419, I understand (though I'd add a small caveat that this isn't really a site for open-ended exploration). I was primarily responding to the statement about brightness level equivalences. For some thoughts that might help assess whether it is likely there exists any such set of equivalences, I recommend Yuval's excellent answer. | |
| Jan 5, 2016 at 19:56 | comment | added | zoplonix | I'm not disagreeing with the exponential aspect of it I completely agree and understand. However saying that you don't think there is enough equivalences to make it practical is not the same as having undeniable proof that it is not practical in the first place (don't mathematicians study equivalences for a reason?). That is what I would like to work toward - proving whether or not it is possible. Just because you can't imagine possibilities does that prove that they do not exist or aren't worth investigating? Does it mean we have to halt exploration immediately? | |
| Jan 5, 2016 at 17:06 | comment | added | D.W.♦ | @user1886419, OK. FYI, the number is still exponential even after taking into account "brighness level equivalences". A simple way to see this is to count the number of black-and-white images, and note that this is a lower bound on the number of greyscale images, even after reducing brightness level equivalences. I don't think there will be any useful set of symmetries that reduces the growth rate enough; for any plausible set of symmetries I can imagine, the growth rate is still exponential. Probably the reason you haven't seen any research into that is that it's clearly doomed to fail. | |
| Jan 5, 2016 at 13:25 | comment | added | zoplonix | My premise was that it's reasonably possible to enumerate all such images but you can see in my conclusion how I say even after eliminating the brightness level equivalences (detailed more in the .nb) there are still to many to enumerate at a high resolution. I haven't seen research into eliminating equivalence cases with images. Thats the main point of what I am saying is that there are equivalence cases to be discovered and eliminated. The quip about being reasonably possible to do it is to motivate the discussion but is not intended to be the conclusion | |
| Jan 5, 2016 at 5:47 | history | answered | D.W.♦ | CC BY-SA 3.0 |