It renders pretty fast, actually. The knock on splats is that they create huge file sizes for video data.
1,685 karma · joined September 17, 2015
Extrinsic calibration (where is my camera in 3D space, and how is it oriented?) being different from intrinsic calibration (how do the pixels in my image map to ray directions relative to the front nodal point of the lens?).
Ergo, I would not bet against teams that understand that objective technical field; AI is not magic, and knowing the fundamentals will always help. Better models, smaller models, faster models; the less you leave for the model to solve in latent space the better you’ll do, I think.
This is often elided by camera systems, that apply a gain to peripheral pixels to correct for this phenomenon. If you understand imaging, you will expect that, and understand why, for example, your wide angle lens displays a lower signal-to-noise ratio for a given illumination value than you might otherwise expect at the edges of the image.
This is a really specific example, but there are dozens. Imaging is its own deep, technical field that is abstracted, and occasionally obscured, by the pinhole model.
Now, I may be biased, because I work in imaging-for-humans, and I've had many a conversations with engineers about why a particular simplification doesn't work for, e.g., filmmakers, but I think that even for purely technical disciplines, understanding the assumptions that go into the pinhole model can be useful. At the margins. Which sometimes matter.
"Cameras, by design, also introduce a level of distortion": distortion is an artifact that is usually minimized in lens design if you are trying to create a rectilinear lens, though you are typically forced to trade it off versus other design criteria (like resolving power, or optical complexity). If you're trying to design a rectilinear lens, you explicitly try to remove any distortion; for "normal" focal lengths (focal length ~= image circle diagonal) you can often get vanishingly low distortion, even with relatively simple lenses. (There was a 19th century lens design that achieved "zero" distortion with just a few elements, and a precisely-chosen aperture position; the name escapes me at the moment.) If you are not designing a rectilinear lens, there are other lens mappings, in which case it's not really proper to describe the effect as distortion, though in technical literature it's often still described this way. (For further reading, check out F-theta vs. F-tan-theta lenses: https://en.wikipedia.org/wiki/Fisheye_lens#Focal_length; https://www.thorlabs.com/newgrouppage9.cfm?objectgroup_id=10....)
"... the proportions of objects closer to the edges of the frame are distorted." This is technically incorrect; objects subtend the same number of pixels in an f*theta fisheye image whether they're at the center of the frame or at the edge, for a given camera-subject distance. The apparent distortion exists because you're viewing it further away than the focal length implies, so the viewer at a "normal" viewing distance is applying a transform to the image that is not distance-preserving. If you get extremely close to the image, and if it were displayed on a curved surface (so the image's angular extent to the viewer matched the camera's FOV), you'd see no distortion. Also, we think of arranging subject matter for, e.g., a portrait, along a plane that's orthogonal to the camera's axis -- but then people closer to the edge are farther from the camera, and would naturally appear smaller. It's a weird artifact of rectilinear lenses that they're actually magnified in the resulting image so people standing in a line are all equal size; to get the same effect with a fisheye lens, you just have people arranged in an arc centered on the camera. (Most portraits are not shot on fisheye lenses though, for very good reasons, so this is more of an edge case.)
Related: "there are no straight lines" -- if you're using a camera to make an image, you're operating in projective space. Whether a line is straight in 3D space, and is then straight in the projective space, is a function of your projection. Rectilinear lenses -- for which the relation f*tan(theta) applies -- straight lines remain straight. But for most (all?) other mappings they do not; for example, f*theta lenses preserve angular extent: an object that subtends a particular angle from the camera's perspective will occupy the same number of pixels regardless of where they appear in the image. This is arguably superior for many computer vision tasks.
... I need to stop here, because this is already getting tedious. For anyone interested in developing a deep intuition for lenses and imaging, I would recommend playing around with a view camera, understanding the Scheimpflug relationship (https://en.wikipedia.org/wiki/Scheimpflug_principle), and really thinking through imaging from first principles if you want to understand these things more deeply.
Or just, you know, use Colmap and don't look back. That works too.
For critical imaging, you still want to calibrate, but the gap between explicit calibration and "implicit calibration" (just using your subject matter) is definitely getting smaller, and quickly.
I will not fault anyone for not believing the certification, but I would personally want to see a more authoritative source than an undated ceramics FAQ. But this is just personal style! Being suspicious of regulatory thresholds is prudent.
(And this FAQ has me wanting to revisit my priors and research this a little more specifically, because its assertions are plausible, but need sources or explanations or dates, for cross-checking. Food-grade epoxy I believe in would be a handy thing to have exist; I do have some certified food grade epoxy in my garage that I haven’t used for anything yet and I’d dislike downgrading it to just “epoxy.”)
I mention this because I was surprised to realize that upon reaching the end of the article, and then of course I had to interrogate why I was surprised. I’m sure this will get some (justifiable) hate, but: I was more sympathetic to the arguments after realizing the author didn’t conform to my stereotype of a mathematically-focused man who undervalued other disciplines.
(I clearly have a blind spot around women who undervalue other disciplines.)
I also seem to recall that the speed of light below c is actually the group velocity, and each individual photon still would move at c. I'm also not entirely sure if photons can be said to exist except at creation and absorption; isn't a photon a phenomenon best described by particle interactions, and moving through free space it's more correctly described as a field? Genuine question, though I somehow doubt I'd understand any good elaboration.
IANAPhysicist, though. I just play with light recreationally.
A finer point: the five-paragraph essay that was popular -- no, required -- in my 90s-era American high school is a great format for creating forgettable text. It's a succinct recipe for making an outline slower to read. Thankfully, we have ChatGPT for that now. But still, real writing can aspire to something more visceral, or raw, real, or just damned interesting.
(For the record, I had excellent English teachers... but the tyranny of the prescribed essay spares no one. No slights, Mr(s). Neely, Vosovic, Dewar, and Green!)
That said, I believe the chain of logic (haven't watched the PBS video yet) is simply that were these fine-tuned constants to take any other value, there wouldn't be intelligent life to observe them. If the values were to be anything outside a narrow range, they would remain unobservable by entities within that hypothetical universe, and because we are making an observation we are implicitly sampling from the distribution of observable values. It's a Bayesian metaphysical argument?
That sounds like it presumes a multiverse, but I don't think you need an infinite number of universes or a god for that to be true... that said, it does purport to explain how fine-tuning doesn't violate certain (metaphysical?) principles of science that call for "naturalness" (which a friend once told me boils down to "all unitless constants should be either 1 or 0 otherwise it's inelegant" or something): the fine structure constant is what it is because otherwise nothing would exist to observe that it was 1/139 or 42 or whatever.
I hope this is even slightly more satisfying to read than it was to write.
But also, the experience is watching something close up is qualitatively different — yes, focus and accommodation, but also relative motions if it’s handheld, and just the perception of size. (Same FOV, but you “know” it’s small and close rather than big and farther away.)
I think Kevin Kelley’s larger point is that our needs are either more plastic than we thought or there are wrinkles in the sorts of things we’re willing to accept, and that’s sort of interesting and fun to think about.
The best business ideas I’ve heard for crypto inevitably don’t need crypto.
The “point” of the book is irrelevant; chapter 21 is disconnected from the main text, set at some remove from the events in the previous chapters, and utterly banal. It’s an extended internal monologue of the most boring sort where the author states his moral.
But the best works of literature, in my opinion, transcend simple morality tales and touch something a bit more universal and hard to define in language. If your point can be summarized in a short essay, all you’ve done is illustrate an essay at length. “This novel could’ve been an email.”
What is captivating about _Clockwork_ is precisely its depiction of violence, its almost sympathetic glorification of the id. The protagonist’s unapologetic thirst for violence is what makes the novel (and movie) interesting. It’s reminding us that violence and aggression are part of the human condition, and part of the reader, too: you may recoil, but you also identify with Alex — it’s a first person account so that’s natural.
The last chapter ruins it though with petty moralizing. The voice of Alex the psychopath is far better than that of Burgess the moralist, and the novel is better for having its terminal essay removed.
(That is not to say the article is wrong, or that recycling is perfect, just that certain facts in isolation don’t really allow you to draw any conclusions.)
Startups without a long history end up in the worst category be default, iirc.