blt

76 ms·

901.
by blt 8y ago
but audio telephony equipment generally assumes the signal is in the human vocal range, right?
902.
by blt 9y ago
I don't know anything about Swift, but I really wish ML had settled on a language with higher performance. Interpreter speed doesn't bottleneck vision stuff but reinforcement learning really suffers.
903.
by blt 9y ago
Good analogy. With gas cars it seems that the engine bay has become so crowded, all the degrees of freedom are used up by making everything fit. Electric drivetrains give an opportunity to totally rethink the topology of the car with automa
904.
by blt 9y ago
Oh there is one YouTube channel I like that is not listed yet: w2aew. Here is very skilled at designing circuits and using the oscilloscope to demonstrate fundamental (not exclusively beginner) ideas.
905.
by blt 9y ago
I'm sure all the information sources I know about have already been mentioned. I will instead share a few simple things I wish somebody had told me: 1. Don't stock up on components at the beginning. Only buy components that you ne
906.
by blt 9y ago
Why is C++ less portable than C? In my experience, non-portable code happens when the interaction with the OS is not well isolated. Since the major OS interfaces are C header files, I don't see why C++ makes it any worse.
907.
by blt 9y ago
True. Their lines of credit even feel somewhat predatory to me, especially wrt tool storage. I am not a mechanic though.
908.
by blt 9y ago
Photoshop is feature-complete, and the features are implemented well. For example, compare the speed of applying a 200px-radius Gaussian blur in Photoshop vs. Gimp. To make an analogy with more hands-on trades: buying Photoshop is like gett
909.
by blt 9y ago
People so often rationalize away a plausible scenario of government surveillance by showing how it would require someone to be an irrational actor -- under a simplified, incomplete set of assumptions about the world. I do not understand why
910.
by blt 9y ago
> Companies operating in Germany have to make sure that data is anonymized and stored in accordance with the law. These data rules are a big reasons why German companies have to stick with their old tools. But every day that we Germans
911.
by blt 9y ago
No. Machine learning is not magic. Machine learning has nothing to say when the objective function is not well understood.
912.
by blt 9y ago
Build quite large test facility with diverse artificial scenes. Pay people to walk/bike/drive around. Use external motion capture, GPS, stationary radars, etc. with additional offboard computation to act as a watchdog and step in
913.
by blt 9y ago
> Isn't there considerable value in having a simple framework that can achieve all those effects, and be shared and communicated easily with others? historically, gfx apis went the opposite direction. fixed function pipeline was rep
914.
by blt 9y ago
lmm the data does not support your claim, see gpm's comment above.
915.
by blt 9y ago
yeah, the deblocked image looks a little softer. Makes sense, the optimization objective encourages it. Sometimes the block boundary aligns with an image boundary. Agree that incorporating a DNN could be a cool solution.
916.
by blt 9y ago
The parser/lexer separation feels especially useless in a Lisp interpreter, where the parsing is trivial.
917.
by blt 9y ago
I don't agree. The typical undergrad CS program doesn't prepare people for AI research, but that's because it dedicates lots of time to discrete/combinatorial stuff and low level systems knowledge. Drop/condense arc
918.
by blt 9y ago
Additionally, this video made particle filters crystal clear to me: https://www.youtube.com/watch?v=aUkBa1zMKv4
919.
by blt 9y ago
Any links to interesting sources about this?
920.
by blt 9y ago
I care about music a lot, but I don't really listen to classical music. I don't actively dislike it, but there are other music genres I like more. Even as a total uneducated outsider, I can tell there is something special about Ba
921.
by blt 9y ago
Fortran, assembler
922.
by blt 9y ago
What is the problem with differentiating a quaternion? E.g. the time derivative of the quaternion w.r.t. a fixed angular velocity is given here: https://fgiesen.wordpress.com/2012/08/24/quaternion-differen...
923.
by blt 9y ago
It is also happening to me. Big CPU usage spikes for no apparent reason. Sucks because I'm trying to reduce dependence on Google.
924.
by blt 9y ago
it would be interesting to try to constrain / regularize this problem such that the solution is palatable. User could answer some questions about ingredient preferences to build approximate set of weights for the LP objective. Also eas
925.
by blt 9y ago
This post made me appreciate being a graduate student a lot. I have many weeks like this!
926.
by blt 9y ago
In my field (robotics, ML) many papers are not interesting or overstate the consequences of the result, but I wouldn't go as far as to say 95% are bullshit.
927.
by blt 9y ago
Is this Verlag related to the academic publisher / rent seeker, Springer Verlag?
928.
by blt 9y ago
cool project! But one question: > A bug that still stands in some Browsers (Chrome and Safari) prevents WebGL from using the Uint8ClampedArray directly. Instead, for these browsers we have to create a Uint8Array view for each array for e
929.
by blt 9y ago
I'm baffled at the amount of debate in this thread. Make is good for tasks that can be expressed as a directed acyclic graph of steps, where the steps' inputs and outputs are files, and steps can be expressed in a few lines of she
930.
by blt 9y ago
oh wow I wish I knew about r_ and c_ a few months ago! I'm still annoyed with numpy for being more clunky than Matlab for linear algebra, but resources like this are good for verifying that I'm doing stuff in a numpy-ic way. Thank

More ›