Silent Speech and Ear-Mountable Devices
OwnerMaximilian Ralph Peter von Liechtensteininventor
Life of this patent
Estimating the term…
Abstract
An ear-mountable device and method for silent speech recognition and software control. The ear-mountable device features a near-infrared light source and optical sensors to detect movements or vibrations of a head-located anatomical surface. One or more processors analyze the sensor signals to detect silent speech events and determine semantic intent. The determined intent is used to parametrize software functions, to control local physical objects, or to facilitate secure payment workflows. The ear-mountable device may be communicatively coupled to a portable or wearable computing device.
Claims20
A system comprising:
- one or more portable or wearable devices comprising:
- an ear-mountable device comprising:
- at least one near-infrared laser diode configured to illuminate a head-located anatomical surface;
- at least one first sensor configured to capture near-infrared light reflected off said head-located anatomical surface;
- and
- at least one second sensor configured as an inward-facing microphone;
- and
- one or more processors of the one or more portable or wearable devices configured to:
- process at least one signal from the at least one first sensor, the at least one second sensor, or both, to:
- detect a silent speech event;
- determine a semantic intent of said silent speech event;
- and
- control, invoke, modify, select, or parameterize a software function with said semantic intent.
A system comprising:
- an ear-mountable device;
- an optical sensor carried by the ear-mountable device,
- the optical sensor configured to:
- emit near-infrared light onto a head-located anatomical surface and to receive reflected light that varies with time according to a movement of said head-located anatomical surface;
- and
- a portable or wearable computing device communicatively coupled to said ear-mountable device, wherein one or more processors of the ear-mountable device, the computing device, or both are configured to:
- process a signal derived from the reflected light to:
- detect a silent speech event;
- determine a semantic intent of said silent speech event;
- and
- control, invoke, modify, select, or parameterize a software function with said semantic intent.
A computer-implemented method comprising:
- obtaining optical information corresponding to silent speech of a user, the optical information comprising at least one of:
- a video stream of a mouth area or face area of the user;
- orinformation derived from reflected light received by an inward-facing optical sensor of an ear-mountable device configured to be worn at least partly in an ear of the user;
- obtaining a body-internal signal from the ear-mountable device,
- the body-internal signal being obtained from at least one of:
- an inward-facing microphone oriented toward an ear canal of the user;
- a laser microphone configured to detect vibration of a head-located anatomical surface;
- a bone-conducting transducer;
- a cartilage-conducting transducer;
- processing the body-internal signal and the optical information jointly to disambiguate a silent speech;
- and
- using said disambiguated silent speech as the basis for:
- controlling a software function.
Dependencies are inferred from the claim text.