Djalil's take on going from seductive theory to applications got me thinking about how difficult it is to convey to newcomers how technology development really works. In fact, this was the reason I drew this figure for the Dorkbot presentation recently as this was also the subject of a discussion on the LinkedIn CS group recently:
The scale on the right points to the maturity of a certain technology before it can be used in some operational fashion specifically in industrial applications. When it comes to compressive sensing related sensors, we are currently in the dip for most of them.
The only technology (MRI) that made the leap did not require a change in hardware.
It is as simple as that. Other ways you will hear about compressed sensing is when it improves not hardware but processes such as in group testing and/or some sort of competitive encoding. This is why you keep on hearing about those but less about the more fundamental new types of sensors.. In effect all the others are so low on the Technology Readiness Level list that it will take some investment from some niche application to grow them up to ubiquity.
- most new sensors being thought up by people featured on this blog will never make it. It's gruesome process: whose heart doesn't sink at the thought of Dirac being inferior to Theora ?
- the math behind CS and beyond, enables the true pirate the confidence to connect different islands and change the way we think about sensors. You don't need to be Lena's prisoner, if she does not look good, maybe you care to much about the wrong model.
Liked this entry ? subscribe to Nuit Blanche's feed, there's more where that came from. You can also subscribe to Nuit Blanche by Email, explore the Big Picture in Compressive Sensing or the Matrix Factorization Jungle and join the conversations on compressive sensing, advanced matrix factorization and calibration issues on Linkedin.