When flying back to HK through Incheon in South Korea. We were flying over a lovely small island which was partially covered by small, fluffy clouds. I took my phone and started to take photos in a fast sequence to make an animation which also happens to give it some depth perception. Ok, now I have a spare moment, so I imported the pictures to Picasa, color-corrected them to remove the grayness caused by the distance and exported them to Gimp in order to make the animation. And, while Gimp was struggling reading the EXIF metadata (I haven't upgraded the poor program in a while..) I got a notification on my Android about the freshly uploaded photos. I opened it and there, in my Google+ albums was sitting a complete animation from all those pictures.
wut
Ok, this won't be so difficult to implement in an algorithm, but still, the idea that they include details like this means something. They must be pretty efficient at Google. And there are some other similar features as well like creating a panorama or a HDR photo.
This little feature blew my mind, because I totally didn't expect that. If I had discovered the unlabeled object recognition feature by accident before reading about it, I would probably first try to wake up from a dream. I have no idea how in the world could they do that.
I wonder what's next and I'm looking forward to it.
