WO2024225007A1 - Hearable device, integrated circuit, and biological signal measurement system - Google Patents

Abstract

The present disclosure relates to a hearable device, an integrated circuit, and a biological signal measurement system that make it possible to acquire a cleaner biological signal. A noise reduction unit executes noise reduction of a biological signal of a user measured by means of a biological sensor, using an output signal from at least one of a microphone and an inertial sensor as a reference signal. The present disclosure can be applied to a hearable device, such as a TWS earphone.

Description

Hearable devices, integrated circuits, and biosignal measurement systems

The present disclosure relates to a hearable device, an integrated circuit, and a biosignal measurement system, and in particular to a hearable device, an integrated circuit, and a biosignal measurement system that enable the acquisition of cleaner biosignals.

In recent years, many different services have been proposed that use biosignal measurement technology around the ear using EEG and PPG sensors built into hearable devices such as earphones and headphones.

Non-Patent Document 1 proposes a technology that uses microphone signals to remove noise from in-ear brain waves.

V. Goverdovsky, W. V. Rosenberg, T. Nakamura, D. Looney, D. J. Sharp, C. Papavassiliou, M. J. Morrell, and D. P. Mandic, “Hearables: Multimodal physiology in-ear sensing ” Scientific Reports, vol.7, Article no. 6948, 2017

However, in actual expected usage conditions for hearable devices, there is a possibility that a lot of noise may be mixed in due to the user's body movements and the external environment. In addition, as exemplified by the technology in Non-Patent Document 1, it is practically difficult to place a microphone directly under the electrode inside the earpiece and use the microphone signal as a reference signal for noise removal.

This disclosure has been made in light of these circumstances, and makes it possible to obtain cleaner biosignals.

The hearable device of the present disclosure is a hearable device that includes a microphone, an inertial sensor, a biosensor that measures a user's biosignal, and a noise reduction unit that performs noise reduction on the biosignal using an output signal from at least one of the microphone and the inertial sensor as a reference signal.

The integrated circuit disclosed herein is an integrated circuit equipped with a noise reduction unit that performs noise reduction on the user's biosignal measured by a biosensor using an output signal from at least one of a microphone and an inertial sensor provided in the hearable device as a reference signal.

The biosignal measurement system disclosed herein includes a hearable device, a noise reduction unit that performs noise reduction on a user's biosignal measured by a biosensor using an output signal from at least one of a microphone and an inertial sensor provided in the hearable device as a reference signal, a feature extraction unit that extracts features from the biosignal after the noise reduction has been performed, and a user state estimation unit that estimates the state of the user based on the extracted features.

In the present disclosure, noise reduction is performed on the user's biosignal measured by the biosensor using an output signal from at least one of a microphone and an inertial sensor provided in the hearable device as a reference signal.

1 is a block diagram showing an example of the configuration of a biological signal measuring system according to the present disclosure. 13 is a flowchart illustrating the flow of a service providing process. FIG. 1 is a block diagram showing a configuration example of a TWS earphone. 1 is a flowchart illustrating the operation of the TWS earphone. 1A and 1B are diagrams illustrating an example of the external configuration of a microphone. FIG. 2 is a diagram showing an example of microphone arrangement. FIG. 2 is a diagram showing an example of microphone arrangement.

Below, we will explain the form for implementing this disclosure (hereinafter, referred to as the embodiment). The explanation will be given in the following order.

1. Prior art and problems thereof 2. Biosignal measurement system and service provision process of the present disclosure 3. Configuration and operation of TWS earphones 4. Appearance of microphone and example of its arrangement 5. Others

1. Conventional technology and its problems
In recent years, various services have been proposed that use biosignal measurement technology around the ear using EEG (electroencephalography) sensors and PPG (photoplethysmography) sensors built into hearable devices such as earphones and headphones.

However, in the actual expected usage conditions of hearable devices, there is a possibility that a lot of noise may be mixed in due to the user's body movements and the external environment. In other words, there are currently challenges in constantly obtaining clean data in everyday life and providing services that use that data.

Furthermore, in devices that need to be small and lightweight, such as TWS (True Wireless Stereo) earphones, there is a particular need to make effective use of resources within the limited space and power consumption. If it were possible to measure biosignals with low power consumption and little noise in a comfortable hearable device, users would be able to receive high-quality services more easily.

Non-Patent Document 1 proposes a technology that uses a microphone signal to remove noise from in-ear brain waves, but the microphone is built into a sponge-like earpiece that is equipped with electrodes, making it difficult to apply to a device designed for listening to music.

If the output signals from the acceleration sensor and microphone built into many earphones could be made multimodal and used as a reference signal, it would be possible to efficiently remove noise caused by various movements in daily life and obtain only clean biosignals. Furthermore, given limited resources such as space and power consumption, it would be desirable to make maximum use of the information obtained from already implemented functions. However, since the microphone built into earphones is designed to capture sounds in the audible range, it is relatively sensitive to high frequencies, but has low sensitivity to low-frequency noise such as body movements, making it difficult to use it as a reference signal for removing noise from biosignals as is.

2. Biological signal measurement system and service provision process according to the present disclosure
The biosignal measurement system disclosed herein is able to make the most of information obtained by functions implemented for other purposes in hearable devices such as earphones and headphones, which have limited resources such as space and power consumption. This makes it possible to efficiently remove noise that is mixed into biosignals due to various actions in daily life, obtain only clean biosignals, and provide services using these to users.

(Configuration of biosignal measurement system)
FIG. 1 is a block diagram showing an example of the configuration of a biosignal measuring system according to the present disclosure.

The biosignal measurement system 10 shown in FIG. 1 is composed of a hearable device 100 and a computer 200.

The hearable device 100 is configured as an earphone or a headphone that is worn on or around the user's ear. When the hearable device 100 is configured as an earphone, the hearable device 100 may be a canal-type earphone or an in-ear-type earphone. The hearable device 100 may also be a TWS earphone.

On the other hand, computer 200 is configured as a smartphone, tablet terminal, PC (Personal Computer), etc. owned by the user. Computer 200 may also be configured as a server on the cloud, etc.

The hearable device 100 can transmit various information to the computer 200 via wireless transmission over a wireless communication network such as BLE (Bluetooth Low Energy) (registered trademark) or BAN (Body Area Network).

The hearable device 100 is configured to include a biosensor 110, an inertial sensor 120, a microphone 130, and an IC (Integrated Circuit) chip 140. Although not shown, the hearable device 100 also includes an acoustic function that outputs music or sound to the user's ears. In addition to the inertial sensor 120 and microphone 130, the hearable device 100 may also include other sensors, such as an impedance sensor.

The biosensor 110 is provided inside a housing that constitutes the exterior of the hearable device 100, or in an earpiece or ear pad when the hearable device 100 is configured as a canal-type earphone. The biosensor 110 measures the biosignal of the user wearing the hearable device 100 via electrodes that contact the skin of the user. For example, the biosensor 110 is configured as an EEG sensor or PPG sensor, and measures the user's EEG signal or PPG signal as the biosignal.

The inertial sensor 120 is provided inside the housing of the hearable device 100. The inertial sensor 120 is configured as an acceleration sensor and a gyro sensor for tracking the user's behavior, and outputs measurement signals that measure six-axis acceleration and angular velocity as output signals to the IC chip 140.

The microphone 130 is provided inside the housing of the hearable device 100. When the hearable device 100 is configured as a canal-type earphone, the microphone 130 may be provided in an earpiece or an ear pad. The microphone 130 may be configured, for example, as a feedback microphone (FB microphone) for noise canceling. The microphone 130 outputs a picked-up sound signal that picks up external sound from the hearable device 100 to the IC chip 140 as an output signal. The low-frequency characteristics of the microphone 130 are at least -15 dB/Pa or more in order to improve sensitivity to low-frequency noise.

The IC chip 140 is configured as an integrated circuit according to the present disclosure, and realizes the function of efficiently removing noise mixed into the biosignal and acquiring only the clean biosignal. The IC chip 140 realizes the noise reduction unit 151 and the feature extraction unit 152 as functional blocks.

The noise reduction unit 151 performs noise reduction on the biosignal measured by the biosensor 110, using the output signal from at least one of the inertial sensor 120 and the microphone 130 as a reference signal for noise reduction. The biosignal after noise reduction is supplied to the feature extraction unit 152. In addition, in the noise reduction, output signals from the other sensors described above may be used as reference signals for noise reduction, if necessary.

The feature extraction unit 152 extracts various features from the biosignal after noise reduction has been performed by the noise reduction unit 151. The extracted features are transmitted to the computer 200 via wireless transmission and are used to estimate the user's state.

Computer 200 is configured to include a CPU (Central Processing Unit) 210.

The CPU 210 implements various programs (algorithms) and executes them to realize various functions.

The CPU 210 realizes a user state estimation unit 221 and an application execution unit 222 as functional blocks.

The user state estimation unit 221 estimates the state of the user wearing the hearable device 100 based on various feature amounts transmitted from the hearable device 100. The user state estimation by the user state estimation unit 221 includes estimation of the user's emotions and stress.

The application execution unit 222 provides various services based on the estimation results by the user state estimation unit 221.

(Service provision process flow)
The flow of a service providing process performed by the biological signal measuring system 10 will be described with reference to the flowchart of FIG.

In step S11, the noise reduction unit 151 of the hearable device 100 acquires the biosignal measured by the biosensor 110.

In step S12, the noise reduction unit 151 performs noise reduction on the biosignal using the output signals of the inertial sensor 120 and the microphone 130 as reference signals. Specifically, by using the output signal of the inertial sensor 120, which has high sensitivity to low-frequency noise, as the reference signal, it is possible to remove low-frequency noise that is mixed into the biosignal. In addition, by using the output signal of the microphone 130, which has high sensitivity to high-frequency noise, as the reference signal, it is possible to remove high-frequency noise that is mixed into the biosignal.

In step S13, the feature extraction unit 152 of the hearable device 100 extracts various features from the biosignal after noise reduction (noise has been removed). The various extracted features are transmitted to the computer 200 by wireless transmission.

In step S14, the user state estimation unit 221 of the computer 200 estimates the user's state based on the various features transmitted from the hearable device 100.

In step S15, the application execution unit 222 of the computer 200 provides various services based on the estimated user state.

By using the above process, the output signals from the inertial sensor 120 and the microphone 130 as reference signals, it is possible to efficiently remove noise that is mixed into the biosignals due to various movements in daily life. As a result, it is possible to obtain cleaner biosignals, which in turn makes it possible to provide high-quality services.

3. Configuration and operation of TWS earphones
Here, the configuration and operation of a TWS earphone, which is one embodiment of the hearable device 100 of the present disclosure, will be described.

(TWS earphone configuration)
FIG. 3 is a block diagram showing an example of the configuration of a TWS earphone.

The TWS earphone 300 shown in FIG. 3 is configured to include an EEG sensor 310, an IMU 320, a microphone 330, a body movement context estimation unit 351, a noise reduction unit 352, a signal quality estimation unit 353, and a feature extraction unit 354.

The EEG sensor 310 corresponds to the biosensor 110 in FIG. 1 and measures EEG signals in the user's ear, for example, via an earpiece-type electrode.

The IMU (Inertial Measurement Unit) 320 corresponds to the inertial sensor 120 in FIG. 1, and outputs an acceleration signal that measures the inertial movement of the user as an output signal to the body movement context estimation unit 351 and the noise reduction unit 352. The IMU 320 may be a sensor implemented for tracking the user's behavior, or may be a sensor with desired characteristics that is provided at a desired position for noise reduction separately from the above.

The microphone 330 corresponds to the microphone 130 in FIG. 1, and outputs a picked-up sound signal that picks up external sounds from the TWS earphone 300 as an output signal to the body movement context estimation unit 351 and the noise reduction unit 352. The microphone 330 may be an FB microphone implemented for noise cancellation, or may be a microphone with desired characteristics that is provided separately at a desired position for noise reduction. It is desirable for the microphone 330 to have excellent sensitivity in the low range (especially the frequency band (approximately 1 to 40 Hz) related to the biosignal (EEG signal) of interest).

The body movement context estimation unit 351, the noise reduction unit 352, the signal quality estimation unit 353, and the feature extraction unit 354 can be realized by the IC chip 140 in FIG. 1.

The body movement context estimation unit 351 uses the trained model M1 based on the EEG signal from the EEG sensor 310 and the output signals from the IMU 320 and microphone 330 to estimate a body movement context that classifies the type of body movement of the user. Body movement contexts include, for example, walking, running, speaking, chewing, and resting. Other functions of the TWS earphone 300 or external information obtained from the smartphone (computer 200) that is the communication partner may be used to estimate the body movement context.

The noise reduction unit 352 uses the trained model M2 to perform noise reduction according to the body movement context estimated by the body movement context estimation unit 351. The EEG signal after noise reduction is supplied to the signal quality estimation unit 353.

The signal quality estimation unit 353 estimates the signal quality of the EEG signal after noise reduction has been performed, assigns a label representing the estimated signal quality to the EEG signal, and supplies it to the feature extraction unit 354.

The feature extraction unit 354 extracts features from the EEG signal supplied from the signal quality estimation unit 353, and assigns a reliability to the extracted features based on the estimated signal quality (label). The features to which the reliability has been assigned are transmitted wirelessly to the computer 200 (Figure 1) and used to estimate the user's state.

(TWS earphones in action)
The operation of the TWS earphone 300 will now be described with reference to the flow chart of FIG.

In step S31, the body movement context estimation unit 351 and the noise reduction unit 352 acquire the EEG signal measured by the EEG sensor 310, and the output signals of the IMU 320 and the microphone 330. The output signals of the IMU 320 and the microphone 330 are assumed to be time-synchronized with the EEG signal. The sampling rate of the output signals of the IMU 320 and the microphone 330 is set to be equal to or higher than that of the EEG signal (preferably 64 Hz or higher).

In step S32, the body movement context estimation unit 351 estimates the user's body movement context based on the EEG signal and the output signal. For example, the user's body movement is classified into body movement contexts such as walking, running, speaking, chewing, and resting based on myoelectric noise mixed into the EEG signal, trunk movement recognized from inertial motion measured by the IMU 320, and housing vibration and voice information input to the microphone 330. The user's body movement may be classified into other body movement contexts such as sleeping as necessary, or may be classified based on the user's situation obtained as the above-mentioned external information (for example, operating the TWS earphone 300, speaking, etc.).

In step S33, the noise reduction unit 352 performs noise reduction by selecting an algorithm for performing noise reduction according to the body movement context estimated by the body movement context estimation unit 351. At this time, the noise reduction unit 352 changes the type of output signal used as a reference signal or changes the weighting of the output signal used as a reference signal according to the estimated body movement context.

For example, in the case of large, gentle body movements such as walking or running, noise reduction is performed using an adaptive filter that uses the output signal from the IMU 320. In addition, in the case of body movements in which relatively high-frequency noise is dominant, such as speaking or chewing, noise reduction is performed using an adaptive filter that uses the output signal from the microphone 330 or by using multivariate EMD (Empirical Mode Decomposition) analysis.

The type of algorithm selected according to the body movement context is not limited to the above, and a DNN (Deep Neural Network) model according to the body movement context may be selected and applied. Noise reduction may also be performed by a single DNN model that also includes estimation of the body movement context. These noise reduction algorithms may be updated periodically, such as when updating firmware, or may be updated automatically by the user adjusting parameters, etc.

In order to efficiently execute signal processing using these algorithms, the noise reduction unit 352 may switch the processor (IC) that executes noise reduction depending on the selected algorithm. For example, a control IC may be used for the numerical analysis algorithm, and an AI (Artificial Intelligence) IC may be used for the DNN model. More specifically, an AI IC may be used for the noise reduction algorithm, and a control IC may be used for the feature extraction in the subsequent stage. In addition, in the noise reduction algorithm, an AI IC may be used for an algorithm based on a DNN model, and a control IC may be used for an algorithm based on an adaptive filter or multivariate EMD analysis. In this way, adaptive selection of an IC depending on the type of calculation makes it possible to speed up calculations and reduce power consumption.

The noise reduction unit 352 may also perform noise reduction using the impedance measurement value of the electrodes of the biosensor (EEG sensor 310) as a reference signal in addition to the output signals of the IMU 320 and microphone 330.

After noise reduction has been performed in this manner, in step S34, the signal quality estimation unit 353 estimates the signal quality of the EEG signal after noise reduction, and assigns a label representing that signal quality to the EEG signal.

In step S35, the feature extraction unit 354 extracts features from the labeled EEG signal. The extracted features are assigned a reliability based on the label assigned to the EEG signal, and are transmitted to the computer 200 via wireless transmission.

In this way, by performing wireless transmission after feature extraction, it is possible to reduce the amount of information sent to computer 200 in wireless transmission that has limitations on transmission bandwidth and power consumption. In addition, by assigning a reliability to the feature, it is possible to improve the accuracy of the user's emotion estimation and stress estimation in computer 200.

In the computer 200, the user state estimation unit 221 estimates the user's emotions using an emotion estimation algorithm and the user's level of mental fatigue using a stress estimation algorithm based on the feature amount and its reliability transmitted from the TWS earphone 300. The application execution unit 222 processes the estimation results of the user state estimation unit 221 and presents them to the user in the form of images, sounds, etc. via a UI (User Interface).

The above process performs noise reduction according to the user's body movement context, making it possible to efficiently remove noise that is mixed into biosignals due to various movements in daily life, resulting in the acquisition of cleaner biosignals.

In the above, in consideration of the limitations on transmission bandwidth and power consumption in wireless transmission, wireless transmission is performed after feature extraction. Not limited to this, in the future, if the processing power of the IC chip on the edge terminal (TWS earphone 300) side and the RAM (Random Access Memory) capacity improve, the processing after feature extraction up to estimation of the user's state may be performed on the edge terminal side. Also, if wireless transmission technology improves, the EEG signal and output signal (reference signal) may all be transmitted from the edge terminal to the CPU (computer 200) side, and the processing after estimation of body movement context and noise reduction may be performed on the CPU side.

<4. Microphone appearance and placement examples>
Here, the appearance of the microphone 330 included in the TWS earphone 300 and an example of its arrangement will be described.

FIG. 5 shows an example of the external configuration of microphone 330.

The microphone 330 is configured, for example, by providing an IC chip or the like in a case formed in the shape of a flat rectangular parallelepiped, and a sound pickup hole 330h is formed in part of the case. The microphone 330 may be a MEMS (Micro Electronics Mechanical System) microphone or an electrolytic capacitor microphone. When the microphone 330 is configured as an FB microphone for noise canceling, noise cancellation is performed based on the external sound picked up from the sound pickup hole 330h.

FIG. 6 shows an example of the placement of the microphone 330 when the TWS earphone 300 is a canal-type earphone.

The earphone 300A shown in Figures 6A and 6B is composed of a housing 411 and an earpiece 412.

The housing 411 contains a driver unit 421 that converts audio signals into sound. The sound emitted from the driver unit 421 is delivered to the user's ear through a sound guide tube 422. The earpiece 412 is removably attached to the tip of the sound guide tube 422, and has the function of transmitting the sound emitted from the sound guide tube 422 to the ear without distortion while sealing the user's ear and blocking out surrounding sounds. In the earphone 300A of this embodiment, the earpiece 412 also functions as an electrode for the EEG sensor 310.

In the canal-type earphone 300A configured in this manner, as shown in FIG. 6A, the microphone 330 may be arranged inside the housing 411 and configured as an FB microphone for noise canceling that picks up sound inside the sound guide tube 422. This arrangement takes into account the acoustic characteristics of noise canceling. In this case, costs and power consumption can be reduced compared to a configuration in which a microphone is provided separately from the FB microphone for noise canceling.

Also, as shown in FIG. 6B, the microphone 330 may be placed inside the sound guide tube 422 and immediately adjacent to the earpiece 412. This placement is specialized for removing noise contained in the EEG signal from the EEG sensor 310.

FIG. 7 shows an example of the placement of the microphone 330 when the TWS earphone 300 is an in-ear type earphone.

The earphone 300B shown in FIG. 7 is composed of a housing 431 and an electrode 432.

The electrode 432 is provided on the housing 431 at a portion that contacts the concha at the entrance of the user's ear, and functions as an electrode of the EEG sensor 310. The electrode 432 may be made of a conductive resin or a metal.

In the inner-ear type earphone 300B configured in this manner, the microphone 330 is simply placed inside the housing 431 and immediately adjacent to the electrode 432. This placement is specialized for removing noise contained in the EEG signal from the EEG sensor 310.

<5. Other>
The above-described embodiment is configured to obtain EEG signals using TWS earphones and provide services based on the EEG signals. The hardware configuration to which the technology according to the present disclosure can be applied is not limited to earphones, and may be headband-type headphones or a head mounted display (HMD). Furthermore, the biosignal handled by the technology according to the present disclosure is not limited to EEG signals, and may be a biosignal measured by a photoplethysmography sensor, an electromyography sensor, an electrooculography sensor, or the like.

Note that the effects described in this specification are merely examples and are not limiting, and other effects may also be present.

Furthermore, the embodiments to which the technology disclosed herein is applied are not limited to the above-described embodiments, and various modifications are possible without departing from the spirit of the technology disclosed herein.

Furthermore, the present disclosure can have the following configurations.
(1)
A microphone and
An inertial sensor;
A biosensor that measures a biosignal of a user;
a noise reduction unit that performs noise reduction on the biological signal using an output signal from at least one of the microphone and the inertial sensor as a reference signal.
(2)
The hearable device according to (1), wherein the microphone is provided inside a housing.
(3)
The hearable device according to (1) or (2), wherein the microphone has low-frequency characteristics of at least -15 dB/Pa or more.
(4)
The method further includes the steps of: extracting a feature from the biological signal after the noise reduction is performed;
The hearable device according to any one of (1) to (3), wherein the feature amount is used to estimate a state of the user.
(5)
The hearable device according to (4), wherein the feature amount is transmitted to a computer that provides a service based on the estimated state of the user.
(6)
The hearable device according to (4) or (5), wherein the estimation of the user's state includes estimation of the user's emotion or estimation of the user's stress.
(7)
A signal quality estimation unit that estimates a signal quality of the biological signal after the noise reduction is performed,
The hearable device according to any one of (4) to (6), wherein the feature extraction unit assigns a reliability based on the estimated signal quality to the feature extracted from the biological signal.
(8)
a body movement context estimation unit configured to estimate a body movement context of the user based on the biological signal and the output signal,
The hearable device according to any one of (1) to (7), wherein the noise reduction unit performs the noise reduction according to the estimated body movement context.
(9)
The hearable device according to (8), wherein the noise reduction unit selects an algorithm for performing the noise reduction in response to the estimated body movement context.
(10)
The hearable device according to (9), wherein the noise reduction unit switches a processor that executes the noise reduction depending on the selected algorithm.
(11)
The hearable device according to any one of (8) to (10), wherein the noise reduction unit changes a type or weighting of the output signal to be used as the reference signal in accordance with the estimated body movement context.
(12)
The hearable device according to any one of (1) to (11), wherein the microphone is provided immediately adjacent to an electrode of the biosensor.
(13)
The hearable device according to any one of (1) to (12), wherein the microphone is a feedback microphone for noise canceling.
(14)
The hearable device according to any one of (1) to (13), wherein the noise reduction unit performs the noise reduction using an impedance measurement value of an electrode of the biosensor as the reference signal in addition to the output signal.
(15)
The hearable device according to any one of (1) to (14), wherein the biosensor includes at least one of an EEG (electroencephalography) sensor and a PPG (photoplethysmography) sensor.
(16)
The hearable device according to any one of (1) to (15), wherein the inertial sensor includes at least one of an acceleration sensor and a gyro sensor.
(17)
The hearable device according to any one of (1) to (16), wherein the microphone is a Micro Electronics Mechanical System (MEMS) microphone or an electrolytic condenser microphone.
(18)
It can be configured as earphones or headphones.
A hearable device according to any one of (1) to (17).
(19)
An integrated circuit comprising: a noise reduction unit that performs noise reduction on a user's biosignal measured by a biosensor using an output signal from at least one of a microphone and an inertial sensor provided in the hearable device as a reference signal.
(20)
A hearable device,
a noise reduction unit that performs noise reduction on a biosignal of a user measured by a biosensor using an output signal from at least one of a microphone and an inertial sensor provided in the hearable device as a reference signal;
a feature extraction unit that extracts a feature from the biological signal after the noise reduction is performed;
and a user state estimation unit that estimates a state of the user based on the extracted feature amount.

10 Biosignal Measurement System, 100 Hearable Device, 110 Biosensor, 120 Inertial Sensor, 130 Microphone, 140 IC Chip, 151 Noise Reduction Unit, 152 Feature Extraction Unit, 200 Computer, 210 CPU, 221 User State Estimation Unit, 222 Application Execution Unit, 300 TWS Earphone, 310 EEG Sensor, 320 IMU, 330 Microphone, 351 Body Movement Context Estimation Unit, 352 Noise Reduction Unit, 353 Signal Quality Estimation Unit, 354 Feature Extraction Unit

Claims (20)

A microphone and
An inertial sensor;
A biosensor that measures a biosignal of a user;
a noise reduction unit that performs noise reduction on the biological signal using an output signal from at least one of the microphone and the inertial sensor as a reference signal. The hearable device according to claim 1 , wherein the microphone is provided inside a housing. The hearable device according to claim 1 , wherein the microphone has low-frequency characteristics of at least −15 dB/Pa or more. The method further includes the steps of: extracting a feature from the biological signal after the noise reduction is performed;
The hearable device according to claim 1 , wherein the feature amount is used to estimate a state of the user. The hearable device according to claim 4 , wherein the feature amount is transmitted to a computer that provides a service based on the estimated state of the user. The hearable device according to claim 4 , wherein the estimation of the user's state includes estimation of the user's emotion or estimation of the user's stress. A signal quality estimation unit that estimates a signal quality of the biological signal after the noise reduction is performed,
The hearable device according to claim 4 , wherein the feature extracting unit assigns a reliability based on the estimated signal quality to the feature extracted from the biological signal. a body movement context estimation unit configured to estimate a body movement context of the user based on the biological signal and the output signal,
The hearable device according to claim 1 , wherein the noise reduction unit performs the noise reduction in accordance with the estimated body movement context. The hearable device of claim 8 , wherein the noise reduction unit selects an algorithm for performing the noise reduction depending on the estimated body movement context. The hearable device according to claim 9 , wherein the noise reduction unit switches a processor that executes the noise reduction depending on the selected algorithm. The hearable device according to claim 8 , wherein the noise reduction unit changes a type or a weighting of the output signal used as the reference signal, depending on the estimated body movement context. The hearable device according to claim 1 , wherein the microphone is provided in close proximity to an electrode of the biosensor. The hearable device of claim 1 , wherein the microphone is a feedback microphone for noise cancellation. The hearable device according to claim 1 , wherein the noise reduction unit performs the noise reduction using an impedance measurement value of an electrode of the biosensor as the reference signal in addition to the output signal. The hearable device according to claim 1 , wherein the biosensor includes at least one of an electroencephalography (EEG) sensor and a photoplethysmography (PPG) sensor. The hearable device according to claim 1 , wherein the inertial sensor includes at least one of an acceleration sensor and a gyro sensor. The hearable device according to claim 1 , wherein the microphone is a Micro Electronics Mechanical System (MEMS) microphone or an electrolytic condenser microphone. It can be configured as earphones or headphones.
The hearable device according to claim 1 . An integrated circuit comprising: a noise reduction unit that performs noise reduction on a user's biosignal measured by a biosensor using an output signal from at least one of a microphone and an inertial sensor provided in the hearable device as a reference signal. A hearable device,
a noise reduction unit that performs noise reduction on a biosignal of a user measured by a biosensor using an output signal from at least one of a microphone and an inertial sensor provided in the hearable device as a reference signal;
a feature extraction unit that extracts a feature from the biological signal after the noise reduction is performed;
and a user state estimation unit that estimates a state of the user based on the extracted feature amount.

Images