Saltar al contenido

About facial recognition and privacy

reconocimiento facial nsa

The national security agency of the United States or NSA has been in fashion for some time, there is a general feeling that they are being used on us all the time and they know how many tablespoons of sugar we put in coffee every morning.

The news of the last hours is that to identify people they collect millions of personal photos every day that they intercept in emails, text messages, social networks, videoconferences and other forms of communication. The bomb was dropped by the NY Times based on documents released by Edward Snowden, a former agency employee.

This reminded me of a lecture by researcher Alessandro Acquisti at TED on facial recognition where he shares the results of an experiment conducted at Carnegie Mellon University where he works.

What they did was create facial recognition software that works with the computing power of the cloud and uses millions of photographs obtained from Facebook as a database.

As data, he mentions that in 2010 more than 2,500 million photos were uploaded to the social network per month and the majority of them tagged.

The experiment was very simple, they randomly took pictures of students who wanted to participate and uploaded them to their cloud recognition system to obtain the person's name and they did it in 1 out of 3 attempts.

They even created a mobile application that allows you to take the photo, upload it to process it and thus obtain the information of the individual that was shown on the same image. As if it were augmented reality, where you simply point your mobile towards a face and you already know who it is.

The app by the way is not publicly available. In the talk he also comments on other very interesting experiments that were carried out around the information that we share on the internet and how it can be used for better or for worse. It is chilling.

Now, if from the laboratories of a university they can do something like this, imagine what a security agency like the one in the United States can do that has no budget or technology limits. Not only do they know how many tablespoons you put into the coffee, they also know the temperature at which you like to drink it and the marks of the sugar, the coffee and the cup itself that you don't even know what it is.