Despite this technology, Siri can recognize users’ head and mouth movements and receive their commands.
Future Apple devices will likely have lip-reading capabilities so that users can activate Siri without the need for a microphone.
Details of a new patent filed by Apple, titled “Keyword Recognition Using Motion Sensors,” have been revealed, and it mentions providing additional aspects to Siri, such as detecting mouth and head movements, that could increase the accuracy of the smart assistant.
Apple’s patent description states:
“[Data] is received from a motion sensor, for example, recording the user’s motion while uttering a speech input. It is then determined whether a portion of the motion data matches the reference data for a set of one or more words.”
Development of lip-reading functionality for Apple’s Siri
Apple further pointed out that the presence of this motion detection system eliminates the need for microphones that the device uses to constantly listen to users and hear “Siri” or “Hey, Siri” commands:
“[Continuously] recognizing and processing voice data requires processing power and capacity, even when the user is not actively using voice control. When the user speaks, their mouth, face, head, and neck move and vibrate. “Motion sensors such as accelerometers and gyroscopes can detect these movements while consuming relatively less energy than audio sensors such as microphones.”
As Apple explained in its patent, this technology uses a motion detection sensor and the camera does not play much of a role in it. Therefore, in addition to the iPhone, there is a possibility of using it in the AirPod as well as the Vision Pro mixed reality headset.
Apple’s patent has been awarded to two inventors, including Siri engineer Madhu Chintakunta. Chintakunta’s previous project for Apple was a patent that would allow Siri to automatically manage users’ appointments and calls.