Apple’s published patent hints at new voice biometrics feature for Siri
Apple’s recently published U.S. patent reveals a new Apple’s published patent hints at new voice biometrics feature for Siri that would allow only recognize the voice of the device owner, according to a report by Patently Apple.
Originally filed in Q2 2016, the patented technology would ensure that users can only access Siri and the device associated with the digital assistant if the system confirms that the voice belongs to the owner.
Up until now digital assistants have solely been responsive to the speech itself, and not to the individual speaker, which is why anyone is able to use the digital assistant to perform various functions on the device.
In addition, a user may provide a spoken user request to the digital assistant associated with the device, only to unintentionally prompt several other nearby devices to execute the request.
Apple’s patent application describes a method to train Siri to know the voice of the device owner before responding to a command.
In the future, device owners will only be able to access Siri if they enter the correct passcode, scan their fingerprint or be recognized via face recognition. In other words, unless the device recognizes the user on multiple security levels, Siri will not respond to commands or requests.
Previously reported, Apple patented display technology capable of analyzing a user’s fingerprint without a dedicated sensor, a feature rumored to be incorporated in the upcoming iPhone 8 model.