Home » News » Google Working on ‘Face Match’ Authentication Feature for Assistant

Google Working on ‘Face Match’ Authentication Feature for Assistant

Face unlock is easily one of the requested features ever since Apple’s Face ID release on the scene. However, implementation of this tech without the dedicated hardware like Apple means facial recognition can easily be bypassed with just a photo of the person.


The Android’s own Trusted Face system has been prone to such failures in the past, but it now seems Google is working on another facial recognition technique, but this time it’s integrated with Google Assistant. It’s yet to be seen whether this will be used to unlock the phone.

The code of the app’s latest version contains precise information about a new Google Assistant feature called ‘Face Match’, which as the name suggests, appears to be a facial recognition feature of some kind.

As per the Apk teardown of the app’s latest build carried out by XDA-Developers, the new feature is mentioned as a setting in the code of the Google Assistant. Moreover, the code states that the Face Match feature will receive metadata in the form of videos as well as images. Also, there’s a separate page for ‘face Match’ enrollment in the Google Assistant settings directory.

The app’s code ties the ‘Face Match’ feature with an AvocadoOnboarding class. Avocado is the codename for a newly spotted feature in the Assistant settings section, and it requires permission from the user to access the device’s camera. Furthermore, the code revealed that Avocado is dependent on ‘face detector dependencies’ and a ‘happiness’ index which therefore refers to a feature in Google’s Machine Learning API Kit for Face Detection.

By this time, it is quite clear that Google is working on a new iteration of ‘Trusted Faces’ and might replace it with ‘Face Match’ that will work side by side with the Google Assistant. Hopefully, we will be hearing from Google anytime soon concerning this development.

Gideon Akeni
Observe and Conquer


Please enter your comment!
Please enter your name here