I've tried colorizing photos with this code (result here : https://twitter.com/matsiyatzy/status/706270684209680384) using EC2 instances. It's pretty fast, around 10-15 minutes, though you'll definitely want GPU-instances with CuDNN drivers installed (which can be a bit tedious to set up).
I'm not sure if it's a deep learning framework. Perhaps they are in some ways complementary http://hunch.net/?p=2875548. Who knows, maybe CNTK is one big reduction.
Does anyone know if there is any code available for this paper? Would be extremely interesting to see comparisons of the EM-algorithm mentioned with regular SGD in terms of speed and precision.
There's definitely geometric differences between male and female faces. The most distinct difference is that the male chin tends to be larger, giving the face a more "square" shape, while female faces have smaller chins, giving the face a slight "almond" shape. And I'm sure there's more subtle differences as well.
nice easter egg : following the "treasure hunt" leads to red letters spelling out "APRIL FOOLS".
Along each of the trails leading to the red letters, there are also hidden letters that spell out "MMC-900913" (a reference to last years google maps april fools)
I'm impressed with how nicely this handles the whole A/B testing stack. But the statistics have some potentially important shortcomings -- I posted a github issue about it:
(Full disclosure: I authored the ABBA library mentioned by matsiyatzy, so this is incidentally self-promotional. But I wrote that tool with the hopes of improving the use and interpretation of A/B test statistics in the wider community :)
