An interesting development could make Apple’s voice assistant more efficient.
Today, ChatGPT and similar generative AI services play the role of smart call machine eggs, but there was a time when even voice-assisted digital assistants were a miracle.
One of their most famous representatives is Apple’s Siri, which will celebrate its 12th anniversary in October, and now a recently published patent entry shows what the future holds for it.
The solution, registered with the US Patent and Trademark Office (USPTO), is titled “Keyword Detection with Motion Detection” and details how Siri can receive instructions by sensing the physical effects that accompany speech, such as mouth movement. The essence of the concept, reminiscent of lip reading, is that the assistant also monitors the data of motion sensors placed in the device and compares them with patterns associated with previous voice commands.
One would rightly think that this serves to clarify the perception, since even today poor Siri often reaches to the side when pressing the Hey Siri command, but Apple does not mean it. Based on the documentation, motion keyword detection would actually be a good thing, since we don’t need to keep the microphone on all the time.
“Continuously receiving and processing voice data consumes energy and processing power even if the user is not actively using voice control.”
writes Apple, which would fix this by activating the microphone only if signals from motion sensors indicate that the user has given a command to Siri.
“When the user speaks, their mouth, face, head and neck also move and vibrate. Motion sensors such as accelerometers and gyroscopes can detect these movements at a relatively low power consumption compared to sound sensors such as microphones.”
Unfortunately, the datasheet doesn’t reveal which devices Apple will use the solution for, but the AirPods Pro also have a gyroscope and acceleration sensor, so size shouldn’t be a problem. On the other hand, it doesn’t hurt to point out that filing a patent doesn’t guarantee the realization of an idea, so it’s easy for a lip-reading Siri to be stuck on the design desk forever.