ryandamm

11 ms·

91.
by ryandamm 2y ago
This is just describing SLAM ( https://en.wikipedia.org/wiki/Simultaneous_localization_and_... it's a deep field and there are many techniques and tools. Extrinsic calibration (where is my camera in 3D space, and h
92.
by ryandamm 2y ago
There is a series of comments further down in the discussion thread.
93.
by ryandamm 2y ago
I think we have objective information about how protective geometry works. 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, sm
94.
by ryandamm 2y ago
A specific example of where it's useful to know the underlying mechanics: for very wide angle lenses, you will typically get brightness falloff at the edges due to a phenomenon called the cos^4 phenomenon ( https://nvlpubs.ni
95.
by ryandamm 2y ago
You are, of course, absolutely right. I nonetheless think it may have value for some people to understand where the deviations from that simplifying assumption lie, and at least understand the stakes. Now, I may be biased, because I work in
96.
by ryandamm 2y ago
If calibration is a temporary problem, it's because the people building those AI models understand imaging really deeply, and engage with some of the ideas in the linked post. To put it another way: if you want to build AI models that
97.
by ryandamm 2y ago
As a camera nerd, I have to weigh in with some minor annoyances: there are errors in this article that reflect common misconceptions or elisions that computer vision engineers often believe in. Let me debunk a few... "Cameras, by desig
98.
by ryandamm 2y ago
I agree, and I don't think this is a particularly hot take; if you've been working with camera systems, you've seen calibration without explicit calibration routines get better and better over the past few years. For critical
99.
by ryandamm 2y ago
But... can’t it just fill a queue? Don’t get me wrong, this still feels like the likeliest explanation, it just raises more questions.
100.
by ryandamm 2y ago
Certified food grade epoxies are used in industry all the time; in fact, some wine is aged in epoxy-coated concrete tanks, and the inside of soda cans these days includes a thin layer of food-grade epoxy. I will not fault anyone for not bel
101.
by ryandamm 2y ago
Written by a woman, by the way. 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
102.
by ryandamm 2y ago
Hoping a physicist can correct me here, but... I believe index of refraction is a function of the photon being absorbed and reemitted by the electrons in the dielectric material, so it's no longer correct to think of a photon moving at
103.
by ryandamm 3y ago
I had the same reaction to Rothko. Doesn’t make sense until you experience it in person; the color and scale creates a totally different experience than what I was able to imagine looking at photos.
104.
by ryandamm 3y ago
Twitch?
105.
by ryandamm 3y ago
I wish I could vote this comment up 100 times. I feel like the writing taught in high school set me back 10 years; it took sustained contact with actual literature while in college just to recover somewhat. A finer point: the five-paragraph
106.
by ryandamm 3y ago
It’s patenting compressing the raw photo site values before debayering, which saves space and in theory allows for better, non-realtime debayering algorithms (more relevant when the patent was filed).
107.
by ryandamm 3y ago
Previously: RED Sues Nikon for patent infringement. https://news.ycombinator.com/item?id=31532285
108.
by ryandamm 3y ago
In the case of red’s patents, the relevant issue for RAW compression is that the sensor values are compressed before de-bayering. It’s still lossy.
109.
by ryandamm 3y ago
But power doesn’t, right? Happy to be corrected (please, educate me!) but I think this is asymmetric, in practice.
110.
by ryandamm 3y ago
You're not entirely wrong that it's unscientific, I think we're answering metaphysical questions. (It seems like questions of "why" ends up unerringly in either metaphysics or religion at some point.) That said, I b
111.
by ryandamm 3y ago
42” at 8’ is smaller than a typical TV setup in the US, I think? So his point still stands in strict size terms. But also, the experience is watching something close up is qualitatively different — yes, focus and accommodation, but also rel
112.
by ryandamm 3y ago
Even more to the point: what does crypto add in this scenario? Would the company be just as effective doing all that regulatory work, then handling settlement in local currency? (Or just using USD as the “stablecoin”?) The best business id
113.
by ryandamm 3y ago
I strongly disagree, having read the book and seen the movie. 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 exte
114.
by ryandamm 3y ago
I read the article and didn’t see any mention of concentrations. I suspect (though IANAChemist) that finding trace amounts of many hazardous chemicals might be de minimus. Which at the very least renders this article incomplete. (That is no
115.
by ryandamm 3y ago
+1 for anything written by Pettis. Very clear communicator, and his Chinese Financial Markets blog[1] (now hosted at the Carnegie Endowment for Peace) is required reading on the Chinese economy. [1] https://carnegieendowment.org&
116.
by ryandamm 3y ago
Scratch is great. Let me also plug using an LLM to help code a game; I got good mileage on an Arduino project with my son by having him talk to the ChatGPT app and pasting the resulting code into the IDE. Great experience for everyone.
117.
by ryandamm 3y ago
The author is a professor at UC San Diego.
118.
by ryandamm 3y ago
Ended the carried interest deduction would be low hanging fruit if we want to discourage purely rent-seeking financial behavior.
119.
by ryandamm 3y ago
At least in California, businesses pay rates that are partially determined by the number of UI claims the company has paid out. You get a letter-based score that has many different tiers. Startups without a long history end up in the worst
120.
by ryandamm 3y ago
Getting the real world onto holographic displays like virtual and augmented reality headsets. These research directions are all possible foundations for holographic video codecs, basically. Which is exciting!

More ›