Wednesday, 22 November 2017
News Tech

Decoding Apple iPhone X’s FaceID & How Face Recognition Tech Will Soon Make Passwords Extinct


After Apple unveiled the much awaited iPhone X, the entire world has been awestruck with the device’s ability to unlock a smartphone by simply looking at it — through FaceID. And while many might feel that it’s quite revolutionary, the principles of face recognition technology were already in place for quite some time now.

Amidst privacy and security concerns, let’s see how it works.

What is facial recognition technology?

Facial Recognition is the ability of an application or device to detect and recognise a specific face, which can be used for numerous applications. Basically, facial recognition technology allows users to gain access by simply scanning the face. This can be used under various apps, either in way of a convenience, which is seen on Facebook’s automated facial tags or as a security barrier — something that Apple did with their iPhone X.

How does facial recognition work?

Just like fingerprints, each face is unique. And although two people might look alike, there are still vast differences between the two faces that make them different, with factors like skin-tone, texture, curves, contours and numerous other features.

Humans have evolved to recognise other human beings through their facial appearance, and in fact, we have a specific part in our brain called ‘Fusiform Face Area’ responsible for all this. This area calculates the generic pattern of faces and differentiates one from another, just like the placement of eye, the eye colour, the nose shape and placement and so on. 

In the early stage of facial recognition, researchers used the same behaviour and made the software denote the difference in faces through distinct facial features, which was called as a person’s ‘faceprint’.

However, this didn’t work perfectly most times, as for accurate recognition, the faces were needed to be in the same lighting and appearance as they were captured in. Each face is dynamic and ever-changing, with factors like ageing, pose, illumination and emotions (which the researchers call as the A-PIE problem) getting an accurate detection was impossible.

The newer DeepFace facial recognition technology helps to solve this problem with a groundbreaking solution. The artificial-intelligence based Deep Learning algorithm captures a single 2D image and converts it into 3D, and scans the face by simulating the lighting, ageing and pose and emotions, eliminating the A-PIE problem. 

Moreover, the Deep learning algorithm also remembers each time it recognises or fails to recognise a problem, and it prevents itself from repeating its mistakes, getting better after every attempt. Facebook has been using this AI learning algorithm with its automatic facial tagging feature, and has come close to a surprising 97.35 percent accuracy!

Don’t Miss

  • Apple iPhone X and FaceID

    Apple’s facial recognition uses a similar basis for its technology, but its overall application is quite unique. For starters, the hardware. The forehead of the iPhone X contains a plethora of sensors and cameras to accurately recognise a face, which it calls as the ‘True Depth Camera System’.

    Ignore the Proximity Sensor, Ambient Light Sensor, speaker and the microphone. The sensors behind the Face ID magic include the Infrared Camera, Flood illuminator, and the Dot projector and the 7-megapixel selfie shooter.

    DON’T MISS: Apple’s iPhone X Will Support Fast Charging, But Only With An Additional $74 Charger And Cable

    Here’s how Face ID works on the iPhone X 

    Step 1: The first thing that gets active is the Flood illuminator. This lights up your face for other sensors to scan. 

    Step 2: Next, the Infrared camera flashes IR lights on your face detecting its placement, depth and position. Since IR cameras also work in the dark, it isn’t affected by low-light conditions.

    Step 3: Now the Dot projector flashes a whopping 30,000 invisible dots on your face which scans your facial features and creates a unique 3D map of the face. 

    Step 4: After capturing the 3D data, it compares it to the one stored on the iPhone X during the initial setup, and if it matches, you get access to your iPhone X. This all takes place in a matter of seconds. 

    Step 5: Now all you need to do now is slide the screen up, and voila!

    Even if due to some problem you’re unable to get through FaceID, you have the regular pin code as a failsafe, so all the crazy Face ID memes can take a break now!

    But the sensors aren’t the only thing that helps you to unlock the device accurately. iPhone X’s new A11 Bionic chip also has built-in deep learning algorithm, which is capable of conducting over 600 Billion calculations per second. This adapts and learns about the dynamic nature of your face. So even if you have a beard, or you’re wearing glasses or even a hat, iPhone X will recognise you in an instant.  

    Security with FaceID

    With Apple getting access to numerous faces that it scans and recognises, one might be able to think that what Apple could do with such personal data at their disposal. With governments already behind Apple to share confidential fingerprint database from Touch ID, and now with FaceID added to the mix, the entire premise seems to get quite serious.

    Surprisingly, Apple cannot really do anything as according to the Cupertino giant, the FaceID data remains within the phone and not on the Cloud. Moreover, the A11 Bionic chip has its own level of security and encryption on the device for preventing of misuse or tampering of this confidential data.

    So in theory, the government will have to physically breach into each iPhone X to reveal the data, and the iPhone encryption is really impossible to crack which has been clearly evident in the past too.

    Android vs Face ID

    While Apple’s Face ID might have gained quite a buzz in a few days, it was Android that had brought in facial recognition on its 4.1.1 Jelly Bean operating system around four years back.

    Although it just used the front camera and wasn’t really secure or optimised, it was one of the first instances of facial recognition on a smartphone, and the device cost merely around Rs 10,000.

    Samsung did try to take facial recognition to the. next level with its Iris recognition tech on the Galaxy S8 and Note 8, but even this has been proved to be easily duped using a mere photograph. Although we surely can expect Android to buck up its game. However, for now, iPhone X remains the champ in smartphone facial recognition.



    Source link

    Post Comment