1,153 karma · joined October 15, 2009
For example, did they say:
jeune
Jaune
john
je une
Je un
Joan
Je ont
Jean
Je ne
....
Perhaps it's just an unconscious bias on my part. Just for reference, I am fluent in at least two more languages apart from my mother tongue.
http://www.thehindu.com/sci-tech/science/india-successfully-...
Despite their company doing badly technically they are still doing a great job. The UI is as fluid as ever, and the rendering is rock-solid. I wish they survive and continue building this browser.
I still use Opera, and super-apprehensive about this new development. It's still fantastic - the rendering engine change actually benefited Opera, in that it removed so many of the rendering glitches in Opera 12. Some options were lost, but it's still very configurable. I respect Hakum and the team that's still there, but not sure if they'll stick around. It will be a shame if we lose one of the best desktop apps on the planet to spyware/bloatware.
https://groups.google.com/forum/#!search/suraj$20nethack/rec...
What a beautiful game.
It most likely won't :) it's (was) a Nokia.
http://blog.dlib.net/2014/02/dlib-186-released-make-your-own...
It's a brilliant piece of software for a number of things.
http://cs.stanford.edu/people/karpathy/deepimagesent/
This got a lot of press:
http://www.nytimes.com/2014/11/18/science/researchers-announ...
I don't have a revenue target in mind. If I get at least $25 per hour I spend, I'm more than happy. I have no idea what the market rates are though. I am handy with most technology, and have enough equipment at my disposal.
- Get connection request
- Open private tab in firefox
- Temporarily disable noscript blocks for linked-in domains
- Accept the connection
- Close tab, re-enable script blocks
They are way too creepy.
Arcane. Some would call it archaic too :)
To compare how well it works compared to OpenCV's face detector: http://www.youtube.com/watch?v=LsK0hzcEyHI
dlib is also great at picking out facial landmarks: http://blog.dlib.net/2014/08/real-time-face-pose-estimation....
Between dlib and OpenBR, I think almost all of these usecases can be covered without commercial APIs - if you have a large enough training set.
Face Recognition is a whole another matter though.
That's one reason not to use them.
My major gripe is those small round icons. If they had conveyed something useful, say a brand name (L for lenovo, T for Toshiba and so on) that would've been useful. Second, there's no scale. If a scale is tough to fit, at least the plot area could've been split into color-coded regions.