<<[Google] has revealed it is investigating the use of quantum computers to run its next generation of faster applications. (...)
Using 20,000 photographs of street scenes, half of which contained cars and half of which didn't, they trained the algorithm to recognise what cars look like by hand-labelling all the cars with boxes drawn around them.
After that training, the algorithm was set loose on a second set of 20,000 photos, again with half containing cars. It sorted the images with cars from those without faster than an algorithm on a conventional computer could – faster than anything running in a Google data centre today, [Hartmut Neven] says.>>
http://www.newscientist.com/article/dn18272-google-demonstrates-quantum-computer-image-search.html
http://googleresearch.blogspot.com/2009/12/machine-learning-with-quantum.html |
Dammit i was just posting this and the duplicate detector stopped me. You beat me!
Pretty cool stuff though. Neven Vision is doing a lot for them =] |
Cool although I never understand your criterion for positing a news item in the forum (versus posting it in the blog). That's the most exciting thing since the invention of fire ;) |
It is awesome that Google is pursuing this kind of research. It's not yet clear whether this is really quantum computation, though. And therefore, I'm not sure why, as the New Scientist article claims, this computation is faster than one done on conventional computers. Maybe it's because the superconducting hardware that they are using can go very fast due to its low temperature. It's the accuracy of the algorithm, and not the speed, that is emphasized by the recent conference article:
http://www.google.com/googleblogs/pdfs/nips_demoreport_120709_research.pdf
As mentioned in the New Scientist article, many people are skeptical about whether the hardware, supplied by D-Wave, constitutes an actual quantum computer. The Google Research blog post acknowledges this: "Unfortunately, it is not easy to demonstrate that a multi-qubit system such as the D-Wave chip indeed exhibits the desired quantum behavior and experimental physicists from various institutions are still in the process of characterizing the chip." At the end of their paper, they say "Finally, we mention that the experiments presented here were not designed to test the quantumness of the hardware. Results of such tests will be reported elsewhere." So maybe some good results are in the pipeline.
I'm glad to see Google supporting something this ambitious, and I wish them luck. |
