kram1032

Yeah it's a classic divide-and-conquer style algorithm isn't it

(any within the range of course)

yeah, ideally I'd like basically any interval size to be possible. Oh well...

It looks like the rejection sampling version also has the probability gap go to 0 but far more slowly

Really the one big downside to this is how it's locked into powers of 2. The input image has to be power-of-2-dimensioned, and all the intervals it comes up with are also power-of-2-dimensioned. If those two things were relaxed, I think this would be just about perfect

ah nice sim code, I'll try it out, thanks

off by one errors amirite

ok so originally it also should have been k == 0 then?

yeah what a pain to debug haha

thanks!

literally had the same issue with alpha before, duh

OH dangit, right that makes sense

nah, local python script. Rerunning from scratch everytime

changed to b and instantly got 0 again

might still be the problem currently, testing

ah sorry, yes I did, but I changed it back

sorry about the indentations, it just flattens itself everytime :-/

(with your previous version I didn't though)

hmm, seems to not work. I somehow still get zero sized intervals. I'll have to figure out why that happens

How do you get code indentation in chat? The usual stuff seems not to work

that makes more sense, thanks

Yeah might give that a try as well, thanks

OK, then I think I'll introduce b for a minimum scale, but I'll stick to rejecting very skewed rectangles for now. If you have any more insight after more analysis, I'll gladly reconsider though :)

yeah absolutely not an issue

Not entirely sure I like that better than simply rejecting samples though - at least in terms of acceptance rate, it's perfectly fine. No notable impact on time at all

it'd certainly be easy enough to at least restore symmetry by randomly picking which one goes first

that would also further restrict the minimum size of the sample rectangle, right?

The aspect ratio part is different though, as it will rely on two separate samples

ok that makes sense. Easy fix

So that'd amount to simply setting the first line to k==2 and probably changing what it returns then

(plus indentations)

yeah, in principle that'd be fine, but another function I relied on later down the line ended up generating a singular matrix in such cases and trying to invert it

(I'd later change it so the aspect ratio isn't hard-coded in. This was just for testing)

I know I actually asked for this, but it turns out very extreme aspect ratios, as this method tends to give, are, well, too extreme. I have changed the algorithm to simply reject interval pairs if the aspect ratio is too extreme, and try again. Is that a reasonable thing to do or do you think that will distort the other properties? (I still don't want squares, just not as extreme rectangles)

Ok, I really like, that there is no overscan at all here. I guess if the AI can see an area $h w$ and my image has area $H W$ I'd likely want $a_s$ to be $\frac{h w}{H W}$ giving me $\alpha = \frac{1}{2 \sqrt{\frac{H W}{hw}} - 1}$ as a solid starting point. I can make do with powers of 2 for $H$ and $W$ for now, so I'll accept this answer. If somebody found a way to generalize this to arbitrary (rational) aspect ratios, I'd probably switch to that answer instead. For now this will work though. thank you!

Yeah that $\alpha$ was meant as an initial guess. Hyperparameter search would definitely be a more rigorous way to figure out a good value for it. Thanks :)

If I see that right, you are assuming that the minimum coordinate is (1, 1) and the maximum one is (W = 2^w, H = 2^h)? Very minor thing that's gonna vary between languages and such, but most of the time (python included), you'd want to start with (0, 0) and end with (W-1, H-1) - thanks, this looks very promising. I would prefer if the power-of-2 requirement were dropped for a less stringent rational aspect ratio (which is always possible with integer-dimensioned images) Technically I want this to work for floats, but that might be more trouble than it's worth, so your assumption is fine there

cu later then. Thanks!

I'll be there

That's fine

which certainly is unusual

hmm, it's like having a space with a variable number of dimensions

Can you write a proper answer or something?

Not sure what to do with my question though