Breaking News on CCTV

What is CCTV one may ask? CCTV cameras are used for common uses such as surveillance, but they include a unique feature. That unique feature is facial recognition. These cameras are starting to pop up in many cities. It is almost as if someone is always watching you.

Many cities have brought in CCTV cameras to assist in preventing crimes or catching criminals. Studies have shown that these cameras might not be as effective as we think they are. London had set up CCTV cameras all over the city. In 2008, only crime 1 was solved using the 1000 cameras that were set up. These cameras also came at a hefty price, about 800 million dollars over a four year period.

The idea is good, actually really good. Setting up cameras to capture videos of those committing the crimes and then using facial recognition software to identify the rule-breakers. Although, something is going wrong in the execution as the numbers of caught criminals with this software is not increasing. What do you think is causing this? Should there be more cameras? What about privacy concerns for the citizens?

Source: CNN

FaceMate, the Future of Online Dating

Find your FaceMate? Who would have thought that facial recognition software could be used to find your soul mate based upon your facial features? Christina Bloom, the site founder, recently said, “I believe that people are attracted to people with similar facial features”.  Bloom partnered with Face.com to develop an algorithm to match potential matches based on nine unique points on the face. The algorithm focuses on the contour, structure, and proportion of each of these features.

That being said, it begs the question: does it work? Kerri Johnson, an assistant psychology professor at UCLA, said, “there is evidence that general liking improves when people look like you”.  Bloom argues this point with her website because the algorithm does not focus in on coloring, hair, size, eye color, etc.  To continue with her theory, Bloom believes that chemistry is a fundamental building block to relationships and that the rest will follow. Perhaps, Bloom is onto something with her theory and that Findmyfacemate.com will be the future of online dating.

To date, Bloom’s website has a database of 43,000 faces that are waiting to be matched to their soul mate.  Who knows, maybe you are someone’s perfect match?

Sources:

Online Dating

FaceMate.com

Facial Recognition and the New Technique for Detecting Plastic Surgery

A lot of people are getting plastic surgery these days, so "Can plastic surgery hinder the accuracy of facial recognition?" is a relevant question.  There is a lot of money and research going into facial recognition research right now because of its efficiency, and accuracy when it comes to identifying criminals.

It would not be surprising to me to think that someone would go under the knife in order to remain anonymous, or to hide from the government. That is why research is also being done on facial recognition technology with the capability of detecting plastic surgery.  Sparse representation is the new technique being introduced by several computer scientists out of Notre Dame.  This method focuses on facial recognition after plastic surgery.  With this technique, if you have several photos of the same person, before and after plastic surgery, you are able to detect exactly what kind of surgery they have had, and who they are.  The drawback is that there have to be multiple photos in the database of surgery before AND after.  

Even though people with plastic surgery look different, the fact is:  structurally, your face has not changed.  Certain areas and shapes that make up your face will remain constant for the most part.  The sparse representation technique focuses on the nodal points that map out the interior structure of the face to get a match.  These computer scientists are seeing a 78 percent accuracy rate with this technique.  This is not as good as the 92 percent accuracy rate we mentioned in class, but it is a start.  

 If you really want to fool the facial recognition technology, you can.  It would take a lot of work, but it can be done.  A spokesperson from VisionMetric, mentioned in this article that he does not think police will ever be able to uncover disguised criminals with a 100% accuracy rate.  Do you think the sparse representation technique will work?  Can you see criminals going to extreme measures to go undetected by facial recognition software?

References:

New Scientist

What is Facial Recognition?

Look no further, the answer you have been waiting for is finally here! Facial recognition is about to be defined!

Technically speaking, facial recognition is defined as a computer application for automatically identifying or verifying a person from a digital image or a video frame from a video source. This technology is making its course across the nation. Every where you turn, this technology is appearing. 

How does this facial recognition software work, anyways? A face is to be recognized through the use of an image or video clip. Through the use of nodal points, end points that are used to measure variables on a person's face, this software is able to recognize individuals. The human has about eighty nodal points. These variables consist of the width of the nose, depth of the eye sockets, shape of the cheek bones among others. These indicators allow the software to recognize the face that is being photographed.

Once the individual is recognized, many actions can take place. Some use this software to gain access to your interests for marketing plots. Others may use it to hack a system enabling them to get your social security number. This points will be discussed more in depth in a future blog, so don't get too excited!

This software is causing an uprising by those who are concerned with their privacy. Will this advanced software be able to recognize you? 

Sources: Facial Recognition 

Big Brother is Now Watching You at Malls

           Advertising has always been an important avenue businesses use to promote growth.  Almost every consumer loves coupons or special deals they receive from companies selling their favorite products.  Most shoppers like to know what is on sale and for how long, especially if it caters to their needs and wants.  Consumers have seen billboards transform from standard painting on a wall to digital ads.  Some companies have even taken it a step further by having facial recognition kiosks set up in the mall to cater to each of their possible customers.

            The ways in which facial recognition kiosks work is very simple.  Consumers simply go up to the sign, stand in front of it, and a camera scans their face.  It quickly identifies the consumer’s age and gender and shows them specials that are being offered from different companies in the mall.  However, consumers will find that only larger companies are sinking their money into this form of advertising because they believe it offers a form of convenience to the shopper.  For instance, a 40-year-old man might be shown special offers from a steakhouse in the mall and a 20-year-old woman might be shown special offers from particular clothing stores.  There are even some billboards that can recognize a consumer’s exact profile and his or her favorite things, which allows it to tell the consumer exactly where to find deals that would be beneficial.

            Even though I believe this feature is convenient for most shoppers, it does not negate the privacy concerns I have.  Many of the companies that are in the process of trying such advertising methods indicate that there are no privacy concerns since consumers are providing them with their information beforehand.  I simply disagree.  I disagree with having the face of consumers scanned and those images being recorded in a database.  No one can be sure where such information goes or what these companies do with the different information.  I believe this feature puts consumers at risk by too easily lending itself to identity theft.

Android 4.0 Face Unlock App

           Facial recognition is becoming more popular and affordable as time progresses.  One unique facial recognition feature that has been recently released is the Face Unlock application for the Android phone.  The features of this app include the latest technology.  The app allows the phone to scan your facial profile and stores it to the phone. After this process has been completed, you can then look at your phone and it will unlock the device.  This feature allows each person’s device to be even more personal. 

            The argument from many people is that use of the facial recognition software allows the Android phone to be easily “fooled” with the use of another face. Some Android customers believe that anyone can use a similar picture to hack into their phones.  In order to silence the critics, Android has introduced “back-up” recognition software that uses a standard PIN or Patterns.

            I believe this feature is not one that would be beneficial for Android phone users.  Anytime someone uses their face as a source of recognition, they run the risk of allowing themselves to become more vulnerable.  A simple PIN code is indeed more efficient because it is less likely a phone can be hacked, and it decreases the risk of another person easily getting information from your phone by putting their face in front of it to get access.  One primary reason I believe PIN codes should be used is there are always chances of computer glitches occurring in this software.