Apple Face ID vs. OnePlus Face Unlock: This is how OnePlus manages to recognize faces without dedicated sensors

Apple Face ID, OnePlus 5T
The trend in mobiles this year has been quite clear: eliminate all possible frames of the screens. In this race of the mobiles without frames, all the big manufacturers have contributed their grain of sand. However, there is another point in which we wanted to highlight, or at least it will be a trend next year: the facial release.

Apple surprised everyone with the Face ID of the iPhone X and its capabilities. Previously some phones like Samsung had unlocked by iris recognition, but not even close to the accuracy of the iPhone X. Why? Because the technologies used in one and the other are very different. One that at the moment seems to match the Face ID of Apple is OnePlus with its Face Unlock in the new OnePlus 5T.

Hardware technology vs software technology

To recognize a face the iPhone X uses the combination of the whole TrueDepth camera. This system is made up of the FaceTime camera, the 30,000 points infrared sensor and the lighting sensors. As we know, the camera projects 30,000 points on our face and makes a three-dimensional map when authenticating. The captured information passes to the secure enclave processor (SEP) and is analyzed with the stored face as the owner of the device if it is unlocked if it does not match no.

The OnePlus 5T is somewhat different, it does not have any combination of sensors, but simply with the front camera. The rest is all software. This front camera of the OnePlus 5T takes a photograph of the face and compares it with the one stored on the phone if they match it opens the phone if it does not match it does not.

To determine what is a real face and not a photograph and to get to analyze a certain depth, what it does is to analyze the lights and shadows. As OnePlus explains, the reflections on our face, the shadows on our features or even the shadow that the smartphone emits on us determine that it is a real person.

OnePlus 5T

The difference is quite clear, one is limited to the data captured by the camera and from there it deduces the rest of the necessary data while the other collects the data from the infrared camera and sensor to have a more precise analysis of what is catching.

Face Unlock from OnePlus is really fast and is surprising to many because of its precision since it is also complicated to deceive it. But it has a number of clear disadvantages, and the first is that you can forget about using Face Unlock at night or in low light conditions.

At the moment, what OnePlus has done seems to be a temporary solution until it manages to manufacture a system in conditions that allow authentication through facial recognition. It is rumored that the Samsung Galaxy S9 will have a dedicated system as well, but most manufacturers are likely to come up with solutions similar to OnePlus in the coming months.

Who knows, maybe someone gets something decent, like the portrait mode of the Google Pixel made by software and not with a double camera.