111–119 of 119 posts
It is refreshing to see problems being solved by AI that are not LLMs. There are so many day-to-day challenges that we could solve using data, machine learn and some creativity.
It is refreshing to see problems being solved by AI that are not LLMs. There are so many day-to-day challenges that we could solve using data, machine learn and some creativity.
It is refreshing to see problems being solved by AI that are not LLMs. There are so many day-to-day challenges that we could solve using data, machine learn and some creativity.
It saddens me that we're wasting so much of that potential on those stupid stochastic parrots that solve all those non-problems that no one has ever had. It saddens me even more that so many people are absolutely sure that LLMs are "smart", or that they can "think", or even that they're somewhat conscious. And that even if they're not quite that, one more order of magnitude of scale will definitely give us an AGI. Oh that didn't help? Then one more, that will definitely be it.
One real problem that LLMs have solved is that they made natural language processing as a discipline obsolete. They also usually don't suck at summarizing long texts, except when they sometimes do. But that's it, really.
Earlier quoted context omitted.
With regards to the sodium vapor process, an idea has been percolating in the back of my head ever since I saw that video. But I don't really have the budget to try it out. theory: make the mask out of non-visable light illuminate the backing screen in near Infra-Red light. (after a bit of thought I chose near-IR as opposed to near-UV for hopefully obvious reasons) point two cameras at a splitting prism with a near I…
Couldn't this be summarized as the Sodium Vapor technique but with near-IR? Or do I misunderstand something?
The sodium vapor light process was the best tech in the 1950s, Sodium vapor lights were used because they deliver a very pure single wavelength light. But we can do better now. leds natively illuminate with a single wavelength(and we have to put a lot of engineering into them not doing this) and we have cameras that can view frequencies that the eye cannot. put this together and in theory you can do the single frequency illuminated backing sheet mask(green screen) with a frequency that is not visible to the human eye and therefore does not interfere with any of the colors in the final shot.
The sad thing about this is the problems encountered during post from the production team saying "fix it post" during the shoot. I've been on set for green screen shoots where the lighting was not done properly. I watched the gaffer walk across the set taking readings from his meter before saying the lighting was good. I flip on the waveform and told him it was not even (which never goes down well when camera dept te…
Earlier quoted context omitted.
Two problems: - It'll bleed on fast motion. Hair in the wind would just not work. - Incandescent lights are out. You could solve both by having two ghost frames shot very close to the real frame (no need to evenly space the frames, after-all) and using strobing a high powered laser. You'd need very fast sensor or another one optically on the same position.
At some point higher fps solves it. Is 240 fps enough?
The calculation isn't too hard though. The width of a pixel divided by the velocity of the subject on the sensor is the maximum delta(T) between real and ghost frames.
But, again, you dont have to shoot faster. You just have to drop the 180-180 degree phase between a real and ghost frame to be 10-350 degrees. Then your 24 fps is capturing the background as if it were 870 fps
Earlier quoted context omitted.
Debevec tried a version of this: https://arxiv.org/abs/2306.13702
im familiar with this work and specifically they tried replicating the sodium vapor style approach but what worked for poppins level isnt actually good enough for today. Specifically you still end up with light spill that contaminates the foreground, especially for things like the fresnel reflections on the side of a face. the magenta idea was to still do what is basically a color difference key, but increase the col…
A totally fascinating experiment.
It’s fascinating to see the bridge between academic research and industry application here. While Image Matting is a massive research area in Computer Vision, academia often focuses on solving perfect 'benchmarks.' Corridor Crew effectively took that foundational research, like neural unmixing and synthetic training, and adapted it to solve the 'messy' reality of production, like tracking markers and motion blur. It’…
Earlier quoted context omitted.
Yes. But the main issue is in the way they formulate the problem. Their output is always a transparency mask, which of course will never handle distortions.
youd have to train it to also generate and st map of the distortions but creating the ground truth version of that from the synthetic data would add a lot more to render. also its very easy to plausibly fake, its not something humans are good at seeing and knowing its wrong. you can tell its completely missing but accurate vs just distorted in a plausible way is not something most brains are tuned to notice.