apu

53 ms·

181.
by apu 14y ago
Is this a YC company?
182.
by apu 14y ago
On slides 61-69, it seems to suggest that they were storing their photos directly in the database. Is this correct? It seems rather inefficient to do this, rather than storing the images in the filesystem.
183.
Visit global landmarks with photo tours in Google Maps(google-latlong.blogspot.com)
2 points by apu 14y ago|0 comments
184.
by apu 14y ago
For which part? That photos take most of the space? Or that they're virtually un-de-dupable? I'm fairly sure about the former (having seen some private numbers that I'm not allowed to share) -- although you can think through it yourself. Wh
185.
by apu 14y ago
It's nice to see that the PR industry is not suffering in this economic downturn.
186.
Creating a Pitchfork.com API using JSFiddle(lukecod.es)
1 points by apu 14y ago|0 comments
187.
by apu 14y ago
I think tlb perhaps disagrees: http://news.ycombinator.com/item?id=1644285
188.
by apu 14y ago
The number of things that are truly impossible is actually quite small (particularly in the domain of startups). Many things are just conventionally thought of as "impossible" because current circumstances make them seem that way.
189.
by apu 14y ago
FYI: "whet your palate"
190.
by apu 14y ago
When I had just turned 20, I was in charge of business on four continents. What managment class would have given me that opportunity? This is one of those things that still blows my mind -- how people operating in supposedly "non-capitali
191.
by apu 15y ago
Hey, I just wanted to say that I've sent one or more of those links to dozens of people over the past few years. I know "business guys" take a lot of shit on programming-oriented forums like this one, but as a technical guy myself, I found
192.
The Berenstain Bears and the Tyranny of Timeliness(themillions.com)
1 points by apu 15y ago|0 comments
193.
by apu 15y ago
In my (admittedly much shorter) career, I've found that there are essentially 2 axioms to giving good academic talks, and everything else can be derived from them: 1. Your talk is an advertisement for your work, not the work itself. Your on
194.
by apu 15y ago
Full writeup coming soon! (I'll comment here with a link, in case it doesn't make front-page.) Getting the lines is not the hard part, but rather, reasoning about the regions enclosed by them to find the panels. I'm not totally satisfied w
195.
by apu 15y ago
I ended up spending most of the day on it and it's more like 20% with that combo. Fortunately, I found a better way that does get me to 80% or so.
196.
by apu 15y ago
For a more technical description of the best algorithms to do this (and more advanced compensation) automatically, see: http://www.cs.columbia.edu/CAVE/projects/pr_any/ BTW, these algorithms have been well understood and described by the
197.
by apu 15y ago
Unfortunately, nothing in computer vision's ever as trivial as you might imagine, or as it should be.
198.
by apu 15y ago
Oh, it wasn't meant to be snarky or a slam -- I'm honestly just curious what kinds of things require so much code? E.g., is it one or two things that dominate usually in codebases this size, or is it just a LOT of components, each of which
199.
by apu 15y ago
At 11:28 he says, "at last count, there was over a million lines of python running this thing" Having never worked with code-bases larger than ~50kloc, I have a lot of trouble understanding what 1 million lines of code is needed for, especi
200.
by apu 15y ago
It's a relative thing -- wouldn't you agree that C is closer to assembly than it is to, e.g., lisp?
201.
by apu 15y ago
That's my goal too, but often I find that the best way to write good short code is to write a shitty, longer version first. "Write one to throw away" is the quote I remember reading from someone, and I find it applies well to some of the co
202.
by apu 15y ago
I remember someone doing an experiment saving git branches every time he saved in his editor, which would give you some estimate into this: http://www.jukie.net/bart/blog/save-everything-with-git-wip But it doesn't look like he describes
203.
by apu 15y ago
Can't find a reference for this at the moment, but I think PG has since modified the HN ranking code a bit to put interesting stuff at the top. (This is since his last arc dump, so I don't know exactly what it is.) Also, I think the situati
204.
by apu 15y ago
Two important components of lectures are 1. Tying facts and ideas together in a cohesive "story" so that it sticks better in people's minds. Good books are supposed to do this too, but sometimes you don't have the flexibility to choose the
205.
Evelyn Waugh's Stock Responses to Letters(listsofnote.com)
1 points by apu 15y ago|0 comments
206.
by apu 15y ago
For SIFT computation, I recommend VLFeat[1], which there are some python bindings for[2] (although I haven't tried them myself). [1] http://www.vlfeat.org/ [2] http://www.janeriksolem.net/2011/06/another-python-interface...
207.
by apu 15y ago
Perhaps it's just me, but I feel like HN comments overall have gotten significantly more negative in the past few months. I can't recall reading the comments section for any recent startup launch or sale without at least half the commenters
208.
by apu 15y ago
I believe one of the co-founders is Navneet Dalal, who, along with Bill Triggs, developed the "Histograms of Oriented Gradients" (HOG) features that are widely used across all of computer vision†. Thus, I'm guessing that there's some seriou
209.
by apu 15y ago
Not sure if you figured this out, but sama = Sam Altman = ceo of Loopt. I think sama's words on this situation carry a little more weight than your guess.
210.
by apu 15y ago
How does MSR stack up for ML? For computer vision, MSR is far better than Y Research, so I'm surprised it would be very different for ML.

More ›