I am the author of the Javascript solution which you ported from. The code itself was a modification of the one in openCV ( its not a direct port since I did some optimizations to speed things up ). However, the dat file was generated from one of the Cascades trained using the openCV haar training program. To summarize how it works:

1, the algorithm use something called haar filters, its simply box filters with a "White" and "black" area - for each region of the image, you take the sum of the pixels in the "White" area and minus the sum of the pixels in the "black" area. Thus the filters output a simple single number;

- First you generate a gazillion of these filters for different regions of the image with different shapes. such as a 2 horizontal rectangle or 2 vertical ones ( one black and one white ).

- You take the result of the output of the filter to find the threshold which differentiate the faces and non-faces in your training images the best.

- Using all the filters and outputs you have, feed it into an machine learning algorithm called "Adaboost", which attempts to minimize an exponential loss function of the error of classification. These different filters are then assembled with different weights.

- The final structure of the detector is a "detection chain", which is a degenerate decision tree (like a chain) with nodes consisting of the aforementioned filters assembled together. This is how the algorithm achieves its speed, by rejecting non faces early in the detection process. Only when a image region passes all the nodes in the detection chain that its labeld as a "hit";

- After that you scan all regions of the image at all sizes ( brute force ) and then assemble the detection results.

To be honest, the first time that I saw the description of the algorithm, it seemed a little "magical". The underlying reason why these "box filters" work so well is because the human face is well-defined by "boxy" feature such as our eyes, nose, eyebrows, lips..etc. Its a wonderful wonderful application of Machine learning to a specific domain.

This is also why this algorithm has MUCH lower detection rates for cascades trained to detect side view faces, because these "boxy features" that are so well detected by these filters are simply not as prominent in the side view

For more, refer to the Viola and Jones paper , of all the versions out there, I find this the best:

http://lear.inrialpes.fr/people/triggs/student/vj/viola-ijcv...

to be perfectly honest, this detector sucks.... Although its fairly illumination invariant, its not rotation invariant and sucks for side-view faces detection. I've been trying for the longest time to implement the histogram based detector outlined in this paper: http://www.cs.cmu.edu/afs/cs.cmu.edu/user/hws/www/CVPR00.ps

which is also what pittpatt uses for their detector and IMO its a much better detector. However the lack of training images and time has been impeding my progress. It'll be opensourced when I'm done.

Sending webpages as email is broken — let's not do it.

I don't know a way of saying the following without looking like an asshole, considering the occasion.

If you find yourself admiring Naggum's acerbic wit and online presence, that's fine. But please please look for ways to live your life so that when you die, the fact will be discovered not because some irc channel regulars notice your longer-than-usual absence.

Please try very hard to arrange that there are people in your life who you love and who love you, and with whom you're in daily contact. People whose fondest memory of you will be as a spouse, a parent or an intimate friend, and not having been cursed by you to hell and back in flawless and elegant English.

Many of us here can easily rationalize ourselves into writing a five-page missive on a technical topic that will utterly demolish the sack of shit our opponent is, at the same time making the merits of our cause blindingly obvious to anyone with a three-octal-digits IQ. Please balance the time investment for writing that five-page missive against the possibility of spending that time on your family, or, if you're single, and then it's vastly more important yet, on looking for a person of appropriate gender you'd really like to spend time with.

Also, don't be the kind of person who's polite with others in face-to-face conversation, and then appallingly rude to the same person in email or online forums of any kind. You'll find it's easy to rationalize such behavior to yourself by appealing to lofty principles. Don't.

The world is cold, and we'll all be dead in the long run. Please have someone to hug and be stupidly sentimental with, and share some happiness with.

Thanks for listening, and sorry for being an asshole.

This article is just utter trash. I usually stay away from "technology reporting" in the "traditional media," and this is a great reason why.

The author is simply a cheerleader for Apple. I was expecting an actual argument as to why Apple "needs Jobs back now." The author even uses the language used to deride Jobs (Dear Leader, etc) but nonetheless succumbs to the propaganda.

As far as this line from the article:

"I would go farther and argue that not only does Apple need Steve Jobs—the world needs him. In an age when the pace of technology innovation keeps accelerating so much that we often feel overwhelmed, we need someone who can package new technology, make it accessible to us, and deliver it to us in a way that makes it simple, useful, and reliable."

I find less cult-like propaganda in Scientology commericals than this drivel.

Also, not to defend Microsoft but to engage in a FUD campaign by stating that their latest beta OS because it supposedly crashed is totally ridiculous and bias does not begin to describe it.

I think to those not already converted (not "switchers," but those converted to the Apple/Steve Jobs cult) this article is just totally laughable.

Sad fact: Not going to happen anytime soon. Definitely not for Outlook 2010.

They turned to Word for HTML editing/rendering in Outlook 2007 because there was no Internet Explorer team during development. IE7 was released in 2007, but Office 2007 was in development since at least 2003. No IE team means there was no one to send design change requests to. Office doesn't like being at the mercy of other teams, so they break dependencies whenever possible. The lack of an IE team was a perfect example of why this is a successful practice for them.

Word offers a significantly better editing experience and most email clients only render a small subset of standard HTML anyway. Gmail, for example, throws away all headers including CSS. It was a perfectly reasonable cost/benefit trade off for the Outlook team to switch to an editor and renderer they could directly change or influence. There wasn't a big loss of compatability, but there was some. Most HTML email in actual usage renders the same between Outlook 2003 and Outlook 2007 because they would have to be written to the lowest common denominator in order to render correctly in most other mail clients as well.

Outlook 2007 added DRM to emails which, besides being the most annoying thing ever, was a feature highly demanded by customers. This feature is based on Word's DRM.

So, putting on my Microsoft PM hat... let's do a cost benefit analysis of switching Outlook to use the IE8 rendering engine:

Benefits:

1) HTML email authors (who don't pay us anything, by the way) will have a slightly easier time now.

2) HTML email authors will have a significantly easier time in the distant future.

Costs:

1) Second release to break HTML mail compatability in a row.

2) Will need to re-implement DRM. This includes a full battery of security reviews.

3) Will need to re-implement all recent editing improvements.

4) Once again at the mercy of the IE8 team, who already have a huge backlog of work and probably don't care about our DCRs.

shrug Not going to happen in 2010. Probably won't happen in the next release either. When some version of IE supports 100% of all standards and is as fast as a hypothetical future Chrome, such that they have time to deal with the Outlook team, then maaaybe, just maybe they can tackle this... two releases of Office later.... If you're lucky.

using tinyurl to hide your amazon referral url... nice