Hacker Newsnew | past | comments | ask | show | jobs | submitlogin
Was Alan Turing right about the mechanism behind tiger stripes? (simonsfoundation.org)
106 points by nature24 on March 25, 2013 | hide | past | favorite | 24 comments


You can generate your own Turing patterns with OpenCV.

Watching them slowly change and with different parameters can be very mesmerizing:

Video: http://www.youtube.com/watch?v=KKwbDKRKG7k multiscale turing pattern (python+opencv)

Code (Python): https://code.ros.org/trac/opencv/browser/trunk/opencv/sample...

Paper: http://www.jonathanmccabe.com/Cyclic_Symmetric_Multi-Scale_T...


Wow, that is awesome!

I spent the past weekend tying to wrap my head around OpenCV. After 3 days of bagging my head against the wall trying to get template matching to work, I determined that I am not yet smart enough for computer vision. The domain knowledge required seems to be very deep.

So, seeing the cool things people do with the library is inspirational, (but also makes me incredibly jealous).


Have you looked at the book "Learning OpenCV", by Bradski and Kaehler, published by O'Reilly? I haven't done much work with OpenCV, but I have read the entire book and it seemed to be a good introduction. A companion collection of examples is available online, consisting of a set of programs that can be compiled and used as a starting point for further hacking. A Google search on the title, happens to yield a link to the full text.


The O'Reilly book is very good to get a grasp of OpenCV but it is a bit outdated as it doesn't include the C++ API. You can still run through the examples but you might have to use Google a lot since a lot of function signatures have changed since the book was written. However the book is worth reading since it also introduces and explains a lot of computer vision concepts, as well as providing practical examples on how to use them.


Of note, there is a second edition due out in June, its subtitle is "Computer Vision in C++ with the OpenCV Library".


That video also seems to be a pretty good way to demonstrate the limits of lossy video encoding.


These are a little sharper: http://vimeo.com/58321857 http://www.youtube.com/watch?v=sCJjvqjyCCg (720p)

Demo Turing pattern-like fluid in WebGL: http://cake23.de/turing-fluid.html


the first interactive reaction diffusion back in ~1995: http://draves.org/bomb/ among many other CA type algorithms.


No mention of recent developments in biology related to pattern expression or the older idea of heterozygous genetic expression:

The Notch signaling pathway: http://en.wikipedia.org/wiki/Notch_signaling_pathway

Heterozygous genetic expresson: http://en.wikipedia.org/wiki/Calico_cat http://en.wikipedia.org/wiki/Barr_body


Perhaps because those systems aren't chemical gradient systems such as elucidated by Turing?

It's meant to be an interest story/article, not a comprehensive analysis of cell signalling and phenotype display


However, the question is about tiger stripes. And modern cell-signaling analysis has given us the answer. So why no discussion about the answer?


Well, it has and it hasn't. For example tiger stripes don't have anything to do with the NOTCH signalling pathway, and the awesomeness of the x-inactivation in tortoise shell cats also has nothing to do with tiger stripes.

Rather, the Hox genes seem quite tied up with this. The Hox genes are nothing new, they were well characterised when I was doing undergrad genetics in 2006.

This article doesn't directly answer th question of tiger stripes but is trying to tie turing's old paper, which focused on tiger stripes as an example, with new research.

Interestingly it seems like this is all just reaearch which has been pooled together with the particular aim of celebrating turing's 100th last year (see [1]) and little of the research is intrinsically ground breaking or unknown to science, but almost done with the express purpose of celebrating Turing.

[1] http://www.eurekalert.org/pub_releases/2012-02/kcl-spt021712...

By the by, did anyone think that the textile seashell pictured in the article looked like some of the shapes created in iterations of the game of life?


Tiger stripes are an example of biological patterning, as is the Notch pathway impact on neural tissue development and also the fur characteristics in other animals. Barr Bodies, Notch, trisomy, all have major impacts on pattern production.


BBC's "The Secret Life of Chaos" covers some of this from kind of a historical perspective. It's a nice pretty superficial documentary, I enjoyed it: http://topdocumentaryfilms.com/secret-life-chaos/


How Mathematicians Think About Patterns - Professor Ian Stewart FRS

http://www.youtube.com/watch?v=xLkQFC-Z-Vk

there is a section about tiger stripes, leopard spots etc etc


> A protein called fibroblast growth factor (FGF) serves as the activator, and a gene variant dubbed Sonic hedgehog (Shh) acts as an inhibitor.

That brightened my day just a little bit...


I wonder if these patterns are related to male pattern baldness too?


>convicted of engaging in homosexual acts and sentenced to chemical castration. Amid that personal drama, he still found the time to publish a visionary paper

Even in the face of his entire society condemning him to a crippled existence, it's a testament to his fortitude that he continued producing brilliant works, unfettered. We're left to wonder what other insights Turing might have contributed to human progress had he not been stifled and squandered by the people in power in his day.

History is filled with examples of intellect being trampled on by the establishment: Galileo, Socrates, and even now Aaron Swartz. We still haven't grown out of it. We still need to fight it.


Your comparisons don't even make sense. Turing wasn't "trampled on by the establishment." He was part of the establishment. He spent most of his career working for the government, first in the war effort then later in a national research lab. He was the victim of backwards social policy, not someone who was attacked by the establishment because his intellect was a threat.


I enjoy Turing's contributions as much as the next man, but this really isn't an apt comparison. Socrates talked his way into his own grave when he could have easily gotten out of it. Galileo is probably closer, but the church beefed with his work, not the man per se. Meanwhile, Turing worked for the establishment directly and was very definitely embraced as a researcher. His persecution is completely unrelated to his genius; he seems to serve more as a figurehead for examples of homophobic bigotry because of his success and successive fame and recognition, not because his case was unusual.

That said, I'm happy with the situation; I don't see anything wrong with using him for that purpose. I just don't see much of a parallel because his anti-establishment status wasn't tied to his genius, and is in fact completely separate.


Wiki has a slightly different account of the conviction.

He was not sentenced to chemical castration, though that was the result, the treatment was to reduce the libido.

Interestingly Arnold Murray was released on probation.

Source is wiki of course so possibilities of inaccuracies.

http://en.wikipedia.org/wiki/Alan_Turing#Conviction_for_inde...

Also apparently jail sentences for 'gross indecency' were rarer at the time, "(only 174 of the 746 men convicted of 'gross indecency' in 1951 received prison sentences-typically less than six months)"

According to this: http://drvitelli.typepad.com/providentia/2011/04/the-turing-...


what source has no possibility of inaccuracy?


I feel I have to defend your comment here, as t

The conviction for gross indeceny was socially backwards, yes, but the establishment (which Turing was not a part of ) Completely fumbled his genius - just look at the treatment von Braun and others got because their rocketry work was valuable.

Had the establishment recognised the value of Turing I am quite sure records could have easily been expunged.

But they did not aNd there is no reliable indicator we have learnt our lessons


Turing, Galileo, Socrates, Aaron Swartz.

Talk about "one of these things doesn't belong".




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: