1,897 karma · joined July 14, 2008
Embedded development, among other things.
Blog: https://www.keacher.com/
Email: jeff.keacheratgmaildotcom
Consider ice hockey. Most coaches, if down by a goal, will pull the goalie for an extra attacker with about 90 seconds left in the game. It turns out that pulling the goalie significantly earlier, with about 3 minutes left in the game, would be more likely to result in a win:
http://people.stat.sfu.ca/~tim/papers/goalie.pdf
However, coaches are understandably reluctant to pull the goalie so early. When they're down by a goal near the end of the game, they're probably going to lose regardless. The public accepts that. However, if they lose anyway after employing an unorthodox strategy, the critics will be vicious, much as described in the football article.
Stanford is a diverse place. Just because a small number of students are dropping out to start companies doesn't mean that the end of the university is nigh.
http://www.dot.state.mn.us/zippermerge/
More discussion:
I've been selling my photo-deblurring product on the web for about a year now. I've always processed payments through Stripe, and I've always served the entire web site using SSL. The green padlock is in the address bar.
And yet, back in the early days, I got questions from potential customers several times a month asking if my purchase form was secure.
My solution? I added the text "Payments processed securely" and a generic padlock icon near the credit card info part of the purchase form. In the eight months since then, I haven't had a single person ask me whether or not my site was "secure." Silly, but effective.
http://en.wikipedia.org/wiki/Paternal_age_effect
http://well.blogs.nytimes.com/2013/03/29/grandfathers-age-li...
Ah, bootstrapping...
Relevant Cochrane review: http://summaries.cochrane.org/CD008829/flossing-to-reduce-gu...
In short, SmartDeblur is a parameterized non-blind deconvolution tool. It handles only a narrow set of blur kernels: focus errors or camera movements in a straight line. Real-world motion blurs seldom conform to SmartDeblur's limitations. Also, the technology in SmartDeblur is nothing new: the commercial FocusMagic has been doing exactly the same thing as SmartDeblur for almost a decade.
On the other hand, Blurity is a blind deconvolution tool. It handles any type of blur kernel, not just when the camera was out of focus or moved in a straight line. From talking with many actual customers, I've learned that almost all real-world motion blur is non-linear in nature. Blurity is well-suited to those situations.
Of course, I encourage people to try both tools and use whatever works best on the blurs in their images.
- Are you asking about recovery of information that was intentionally blurred? If so, there have been some exciting results in this area in recent years. For example, it turns out that given reasonable priors on the underlying glyphs, it's possible to reconstruct even heavily blurred text. (e.g., http://dheera.net/projects/blur.php ) That doesn't involve deblurring per se, since the goal is simply to find the glyphs (and combinations thereof) with maximum likelihood given the blurred text. Still, it's interesting, and similar techniques have been applied to super-resolution problems lately, where the approach is often called "hallucination."
- Blind deconvolutions often are used by experts. However, it's possible to help things along if a glint exhibiting the blur of interest is visible in the photo. That is: if a bright dot of light is visible in the photo, it will be blurred too, and so in the blurred photo the point of light will look like the point spread function (the blur kernel), and that can then be extracted and used with a non-blind deconvolution approach to recover the latent sharp image.
- Blind deconvolution and super-resolution are related problems, so advances in one might certainly help in the other. The research you cited makes use of image hallucination (mentioned above) as a key part of its super-resolution algorithm.
If the true blur kernel is more complicated -- perhaps a wavy line -- then you probably need a blind deconvolution tool, which this is not (yet?).
If you're interested in blind deconvolution in general, Dr. Levin of MIT put together a nice overview paper a few years ago: http://www.wisdom.weizmann.ac.il/~levina/papers/deconvLevinE...
(Disclaimer: I'm the developer of Blurity, a blind deconvolution product)
A gentleman with whom were were talking mentioned Plimus as another option for countries where bank transfers are common, but they have extremely high fees and low name recognition. Amazon's payment platform could be another option; an older incarnation of Blurity used it.
On the other hand, I've found that the Bing search ads I've run for my software product convert better than my AdWords search ads. Maybe Bing users are more likely to spend money?
http://www.theatlanticwire.com/politics/2012/09/someones-cla...
EDIT: I looked into the two companies mentioned in the article, and as of now, nine years later, Gemesis still hasn't started full production, and Apollo folded (with its assets purchased by another synthetic-diamond company).
After I discovered people were doing that, I was rather surprised that they thought that the trademark serial number would be a valid registration key. I guess people don't read.
Edit: Thanks for the heads-up about the typo. Oops!
On Windows 7 with IE 9, there were no warnings presented when downloading the zip file containing the unsigned installer executable. When the installer was extracted from the archive and run after the download completed, I got the standard Windows security warning about "The publisher could not be verified", which was far less scary that the SmartScreen warning.
So yes, it looks like that could be a viable work-around.
When I was experimenting with the SaaS version of Blurity, I found that many of the people who did eventual make purchases were, firstly, interested in only a single photo; and secondly, satisfied with the most trivial of improvements. I lamented that those people would be just as satisfied with auto-levels and unsharp mask as they were with Blurity, so my friend Tyler threw exactly that simple service together in about 10 hours.
The results? Turns out that people aren't willing to pay for something simple like that after all.