petters

68 ms·

811.
by petters 10y ago
Have you seen the "super-resolution" results with faces and NN? Imagine grainy CCTV footage as evidence in court. A serious image analyst may attempt using super-resolution to obtain the face of a suspect. But a NN could happily i
812.
by petters 10y ago
OK. In my (limited) view of the field these kind of things started with "Super-Resolution From a Single Image" in 2009 but apparently it is older than that. To me personally, it feels weird to call this super-resolution, but I rea
813.
by petters 10y ago
You may be right. It's good to see some reasonable views at least.
814.
by petters 10y ago
When I was still in computer vision, "super-resolution" meant taking multiple images and combining them to really be able to see what was in the scene. Now it often means: take a single image and have the computer guess/make
815.
by petters 10y ago
I'm glad to see Kasparov's views on AI are now reasonable. Eight years or so ago he had the opinion that poker AI would always be impossible (because of bluffing, I think).
816.
by petters 10y ago
Sorry, I misread the theorem earlier. Thanks!
817.
by petters 10y ago
Interesting, I've always known them as Karush–Kuhn–Tucker conditions. Karush had them in his master's thesis like a decade earlier.
818.
by petters 10y ago
That such a function f exists is pretty standard, right? I know we discussed it during my undergrad studies. See e.g. Counterexamples in Analysis. Also, a point is a closed set, so you probably want non-empty interior.
819.
by petters 10y ago
Another measurement put Google at 13% in 2010. https://m.theregister.co.uk/2010/01/29/google_web_server/ Also, by amount of traffic it's another story (YouTube).
820.
by petters 10y ago
Here is someone making it a little more rigorous: https://www.gwern.net/LSD%20microdosing N=1 though.
821.
by petters 10y ago
I think it is quite common to want to spend time with your significant other instead of working. To call that burned out is to exaggerate the problem. I think you should apply to other jobs, just to see what your options are.
822.
by petters 10y ago
They do well for single-person code bases, but velocity scales badly with the number of developers in my experience.
823.
by petters 10y ago
The interviewer could look at the candidates resume and ask an ML question if they want to. The type of position is also available to them.
824.
by petters 10y ago
I just read that; it is even too optimistic in some cases: "Ten percent of those fees went to the researchers who wrote the papers"
825.
by petters 10y ago
What a beautiful response to a completely unnecessary comment! Thanks.
826.
by petters 10y ago
The reason Samsung bothers with software of their own is that they are a little scared of Google. What if Google started charging a lot for the non-free parts of Android?
827.
by petters 10y ago
Would still be an interesting comparison, with some Electron apps using 16% CPU to blink the cursor.
828.
by petters 10y ago
Completely agree about Cmake. The language may be quirky, but it has solved all the hard problems. Often the build files will be trivially short as well.
829.
by petters 10y ago
Now it uses Webassembly and Web workers to run in the background.
830.
by petters 10y ago
I and I think many others would disagree that it would be almost impossible to decode a TV signal.
831.
by petters 10y ago
The fact that animating a cursor at 60Hz requires non-neglible CPU is still a little sad.
832.
by petters 10y ago
This is how it works in Sweden. Most people just approve their "post card." Everything is done online.
833.
by petters 10y ago
That service is the best! However, it requires submission so it does not have as fast compile-run cycle.
834.
by petters 10y ago
Despite the URL, this is not Webassembly. I will try that next when it is more supported.
835.
Show HN: Integer Programming in the Browser(strandmark.net)
52 points by petters 10y ago|5 comments
836.
by petters 10y ago
Perhaps the yearly targets should accumulate, so that if the actual inflation one year was 1%, the target for the next year becomes about 3%.
837.
by petters 10y ago
Funny, I did not even notice that it is unreadable on mobile. I almost subconsciously pinched to get rid of the margins the first thing I did. This problem is extremely prevalent.
838.
by petters 10y ago
Yeah, I play chess sometimes, and I could not care less about how the board is oriented.
839.
Announcing Guetzli: A New Open Source JPEG Encoder(research.googleblog.com)
14 points by petters 10y ago|3 comments
840.
by petters 10y ago
Also, typedef make things like arrays of function pointers much easier to read.

More ›