Thursday, November 19, 2015

Your Life, In Patterns

Do you ever wonder if what you see is the same as others?  Does that cup of coffee look and smell the same to you as it does to the next person?  Is that rose really red?  As humans, most of us have to come to accept the fact that just about everyone perceives reality differently.  Sometimes more, sometimes less, but no two people will see an object in exactly the same way.

Given what I've just said, is it truly a fact that appearance is completely subjective?  The answer is most definitively "No!".  In truth, when the human element is removed from perception, things are definitely no longer subjective.  It couldn't be any other way or else there would be no:

Shazam
Spotify
Soundhound
International criminal database
Microsoft Kinect
Facebook auto-tag

Each one of the items above uses pattern recognition to provide us the ability to identify things that we normally could not.

Let's look at an illustration.

Imagine that you see a picture of a beautiful woman.  You take note of the color of her hair, eye shape, nose, lips, etc.  If asked, you would describe that woman in terms of what you perceived related to her physical features.  Now let's compare how a computer would view that same woman through the filter of a Fourier transform:



Obviously, the Fourier transform graphic looks absolutely nothing like the woman from the perspective of the human eye.  However, from a mathematical point of view the Fourier version represents the complete unique "visual signature" of her.

Computer recognition of images, shapes, and patterns these days has become orders of magnitude more accurate than the human eye.  It's all because computers view images in terms of discrete math.  The human eye would struggle to identify individual coins in a new box of pennies but a computer, using Fourier-based pattern recognition, would see each one as a wholly unique entity.  Good luck if you're a criminal these days trying to avoid detection.  There are cameras located in virtually any populated place you might be and they are *always* evaluating your image.

This science works for audio just as well if not better.  By now, if you have a smart phone odds are that you've used an application like Shazam.  Many times I have been in loud, crowded places where I have heard a song that I wanted to identify.  By using Shazam and other apps, I've simply had to activate my phone and hold it up.  Ten seconds later I have all the data - name, writer, performer, date - on the song.  Pretty amazing technology!

In audition to Fourier transforms, there are other technologies such as audio fingerprinting.  Just like with images, each song is unique when viewed through the lens of mathematical pattern recognition.  While these technologies are very helpful to us when we want to identify a song or sound, they are also frustrating musical artists such as Robin Thicke.

Back in 2013, Thicke (and a few others) wrote and recorded a song called "Blurred Lines".  This is the same song that launched the career of Emily Ratajkowski and ended up being dubbed the "Song of the Summer".  However, there was a little problem.

The heirs and children of the late R&B singer Marvin Gaye sued Thicke claiming that "Blurred Lines" was a knockoff of the song "Got To Give It Up".  Gaye's family won the suit and a huge monetary judgement in large part because the jury was shown evidence, through audio fingerprinting, that there were just too many similarities between the two songs to be coincidence.

So what does this all mean?

Your life and hence the whole world are defined in patterns.  Part of what makes life great is our collective shared experiences of how we all view the world and everything in it.  By looking at our existence through the lens of mathematics and other new, amazing tools, we begin to see just how wonderful and unique we truly are.