This is still more-or-less a significant domain reduction over the human face. If you weren't human you wouldn't be able to identify one face from another. Try identifying faces upside-down sometime in a split second. We are uniquely built to identify very small variations in the human face in upright position. I suspect the front end of a car would be much less difficult. For example, first classify what the make of the car is, then classify color, then classify markings. If you have a database of cars in the race you will find a match with high probability.

http://ask.slashdot.org/comments.pl?sid=2499210&cid=37875118 There are image processing techniques that would work with relatively small identification marks and a sufficiently "good enough" camera. The processing may take 5 minutes but you can cut to commercial and have a human verify it later. The issue is getting the original system accurate enough you are 99 percent sure whoever crossed the line first, crossed it. Facial recognition systems already have this level of accuracy. Just like voting systems however, for fairness you need human review before passing definite judgement. Apply some science and you would have a killer product there. Im surprised there aren't more industries employing at least one academic to make their systems more cutting edge.

Most likely power consumption due to the power needed for each cpu and cooling requirements (which affect the space needed and infrastructure needed to store it all). I don't think an AMD cluster is less reliable over an Intel one in ideal physical conditions for the AMD processors on each. What I mean by that is, sufficient cooling and power supply for the AMD cluster since they are known to be more power hungry and require more cooling.

http://hardware.slashdot.org/comments.pl?sid=2498782&cid=37871308

Sorry, meant higher performance per cluster / cost.

You may be right. Since I don't have concrete evidence here, I will admit that I simply was wrong about saying exclusive. I have been at 2 top tier universities and as such my sample is not large enough to say anything about it in regard to universities in general. I suspect universities buy AMD because its a higher cost/performance per cluster if you don't factor power consumption (nor anything else) into your metric.

Anonymous coward, please don't refer to the Bible as your "tech manual for life", it only discredits you. http://wiki.ironchariots.org/index.php?title=Main_Page

BTW http://hardware.slashdot.org/comments.pl?sid=2498782&cid=37871308

I don't really understand what the fuck that link is supposed to prove other than you have been published. So what? I have been published before as well. Does being published mean you know what you are talking about? Nope. Logic flows from evidence and rational argument, and I see no evidence nor rational argument in your post.

Wait to make the problem impossible due to arbitrary and unrealistic expectations. Approximately 1.5 AMD servers will have the same computing power and cost about the same, except they will have more memory and throughput. Power consumption is a different issue, but then I digress.

I suspect what you mean is that I am a demon because I sided with Teh 3vil |31g |3r0th3r. Technology != evil. Its how its applied that matters. There are legitimate uses for such technology, such as identifying a person who is trying to enter their home, or a person who is trying to access their bank account, or even to identify a person known to make bombs and blow people up who try to board airplanes. Its not necessary to be a luddite to avoid abuses of technology. Fix the system rather than oppress scientific development and innovation. You can band together with like minded people and force laws into existence that protect you from abuse, or you can just stop being apathetic and never allow such abusive laws to pass.

The only reason I suspect its easier is that its easier to convince an admin in control of budget later that "HOLY SHIT! We don't have enough money in our budget to pay for the power our servers need to continue to operate and if we lose the servers we won't attract massive research grants!!! LOL" than, "Hey, why don't we try to spend more and get a higher performance and lower power consumption per CPU?" Of course there are other problems, such as the fact that it looks better when you purchase X number of nodes at Y performance for less than P number of nodes at Y performance when you leave power consumption out and X>P. The University system is just as rife with politics and bureaucratic nonsense as any other place, probably even worse actually.

I worked in Face Recognition as a research assistant for my MS. I was part of a group designing a robust end-to-end system for the purposes of identifying people in a wide variety of conditions. This problem seems easier because you can assume that the cars will be in one location within one time frame, the camera is fixed and everything in your data set will cross this line unless they break down, they will have almost identical 3D structure (unlike the face), and you can expect some identifying marks on the front end that crosses the finish line. As long as you have some sort of identifying marks, you can use standard face/object recognition techniques to identify the vehicle. There are some pretty advanced algorithms out there. The best out there when I was actively working in face recognition was STASM and Pittpatt. Better yet, if you had some unique identifier expected at some part of the vehicle, you can easily make the problem much easier. For example, have each one use an infrared identification tag located in precisely the same spot relative to the build of the vehicle.

Thats a tough one to find a citation for. Essentially, at the three universities I have worked and two of them I have collaborated with they had between 6/10 and 8/10 clusters running AMD processors. Some had NVidia clusters as well. Some had ones running Intel but they were older systems or were in use by departments other than Math/CS. This evidence may be anecdotal, but two of the universities are larger ones with large research budgets. Exclusive was a bit of a exaggeration in hind sight.

Yes, as I said, its cheaper. I did not mention performance per node but I did in a later post.