Interested in licensing this patent?
MTEC can help explore whether this patent might be available for licensing for your application.
Assignees
Member
OmnibudsOmniBuds is a clinical-grade, in-ear sensing platform that transforms everyday earbuds into continuous health, safety, and performance monitors. By combining premium audio, medical-grade biosensing, and on-device AI, OmniBuds delivers high-fidelity physiological insight from the most stable and information-rich measurement site: the ear.
Built on nearly a decade of R&D, OmniBuds is a scalable platform spanning healthcare, enterprise, and defence. From chronic disease management and remote patient monitoring to fatigue, stress, and cognitive load detection in extreme environments, OmniBuds is redefining how we monitor, protect, and save lives.
OmniBuds is a clinical-grade, in-ear sensing platform that transforms everyday earbuds into continuous health, safety, and performance monitors. By combining premium audio, medical-grade biosensing, and on-device AI, OmniBuds delivers high-fidelity physiological insight from the most stable and information-rich measurement site: the ear. Built on nearly a decade of R&D, OmniBuds is a scalable platform spanning healthcare, enterprise, and defence. From chronic disease management and remote patient monitoring to fatigue, stress, and cognitive load detection in extreme environments, OmniBuds is redefining how we monitor, protect, and save lives.
Publication Number
US-12329542-B2
Publication Date
2025-06-17
Expiration Date
Abstract
A method is provided that includes determining a quality of a data portion of an input sensor data stream based, at least in part, on data of a first data type and determining between, at least, generation of two or more streams of a second, different data type including at least one synthesised data stream of the second data type. Determining between generation of two or more streams of a second, different data type is based, at least in part, on the determined quality. The synthesis is based, at least in part, on the data of the first data type. The method further includes causing generation of at least one stream of the second, different data type based, at least in part, on the determination between generation of two or more streams of the second, different data type.
Core Innovation
A method and apparatus are provided that include determining a quality of a data portion of an input sensor data stream based, at least in part, on data of a first data type, determining between, at least, generation of two or more streams of a second, different data type including at least one synthesised data stream of the second data type, wherein the determining between generation of two or more streams of a second, different data type is based, at least in part, on the determined quality and wherein the synthesis is based, at least in part, on the data of the first data type, and causing generation of at least one stream of the second, different data type based, at least in part, on the determination between generation of two or more streams of the second, different data type.
Some electronic devices, such as some mobile devices and/or some wearable devices, are configured to perform sensing, for example sensing in relation to a user to allow determination of one or more bio-signals of the user. It would be desirable to improve sensing by an electronic device.
The disclosure addresses improving, correcting, replacing and/or enhancing bad quality and/or corrupted and/or missing sensor data by determining the quality of portions of input sensor streams using data of a first data type, selecting between generating filtered data streams or synthesised data streams of a second data type based on that quality, and causing generation and, where applicable, replacement of the data portion with the generated second data type stream.
Claims Coverage
The independent claims disclose three main inventive features.
Quality determination of a data portion based on a first data type
Determining a quality of a data portion of an input sensor data stream based, at least in part, on data of a first data type.
Selection between filtered and synthesised second data type streams based on determined quality
Determining, based, at least in part, on the determined quality, which at least one stream of between, at least, generation of two or more streams of a second, different data type to generate, wherein the two or more streams of the second, different data type include at least one synthesised data stream of the second data type and wherein the synthesis is based, at least in part, on the data of the first data type.
Causing generation and replacement of the selected second data type stream
Causing generation of the at least one stream of the second, different data type based, at least in part, on the determination between generation of which at least one stream of two or more streams of the second, different data type to generate, and replacing at least the data portion of the input sensor data stream with a generated stream of the second, different data type.
The independent claims collectively cover determining signal quality using a first data type, selecting between generation of filtered or synthesised second data type streams based on that quality with synthesis conditioned on the first data type, and causing generation and replacement of the input sensor data portion with the generated second data type stream.
Stated Advantages
Improve poor quality sensor signals before they are fed to downstream models and/or applications, thereby reducing the need for increased robustness and complexity at the model and/or application level.
Use correlations between data of a first type (such as motion data) and noise introduced in other sensors to detect and quantify noise and to act as a reference signal for filtering or conditioning synthesis.
Leverage correlations with other sensor data to generate good quality synthetic data when raw sensor data cannot be recovered or is missing.
Provide low latency to supply applications with real time data.
Provide transparency for downstream applications and for a user by delivering seamless, good-quality data even when raw sensor data is poor or missing.
Documented Applications
Sensing to allow determination of one or more bio-signals of a user.
Use in personal devices, mobile devices, and wearable devices, for example a smart watch, a smart ring, a smart ear worn device, and smart glasses.
Personal systems including personal health systems or personal fitness systems.
Use as enabling components of automotive systems.
Use as enabling components of telecommunication systems.
Use as enabling components of electronic systems including consumer electronic products and distributed computing systems.
Media systems for generating or rendering media content including audio, visual and audio visual content and mixed, mediated, virtual and/or augmented reality.
Navigation systems and user interfaces (human machine interfaces).
Networks including cellular, non-cellular, and optical networks; ad-hoc networks; the internet; the internet of things; virtualized networks; and related software and services.
Interested in licensing this patent?