Abstract: The present invention provides an improved method and system for denoising high-frequency ultrasound based on a multipath matching pursuit algorithm. The method includes: acquiring a high-frequency ultrasound detection signal of a to-be-tested sample; constructing a discrete overcomplete dictionary according to the high-frequency ultrasound detection signal, and training the discrete overcomplete dictionary; reconstructing the high-frequency ultrasound detection signal by using a trained dictionary and using a multipath matching pursuit algorithm, and obtaining a global optimal atom; performing interpolation on the global optimal atom, and constructing a consecutive atomic library; and reconstructing the high-frequency ultrasound detection signal in the consecutive atomic library according to a parameter of the global optimal atom, to complete signal denoising.

Abstract: Provide is a wearable device comprising at least two electrodes and a wearable structure. The at least two electrodes are configured to fit against the skin of a human body to collect an electrocardiogram (ECG) signal of the human body, and the wearable structure is configured to carry the at least two electrodes and fit the at least two electrodes to a waist region of the human body. The at least two electrodes are spaced apart on the wearable structure, and disposed on two sides of a mid-sagittal plane of the human body when the human body wears the wearable structure.

Abstract: The present disclosure provides a wearable device. The wearable device comprises at least two electrodes, configured to be attached to a human skin to collect an electrocardiograph (ECG) signal of a human body; a wearable structure, configured to carry the at least two electrodes and attach the at least two electrodes to two sides of a midsagittal plane of the human body; and a processor, configured to determine a heart rate variability (HRV) of the human body based on the ECG signal, and determine a physical state of the human body based at least on the HRV.

Abstract: Disclosed is a signal measuring device including: at least one electrode configured to contact with a human body to collect an electromyographic (EMG) signal of the human body in a target frequency range; an alternating current (AC) excitation source electrically connected to the at least one electrode, the AC excitation source being configured to provide an excitation signal at a first frequency to generate a detection signal reflecting a contact impedance between the at least one electrode and the human body; and a gain circuit configured to provide a gain for the EMG signal and the detection signal. In the present disclosure, by providing the gain circuit, gains are provided for the EMG signal and the detection signal collected so as to improve strengths and qualities of the EMG signal and the detection signal, and enable an accurate monitoring of the signal measuring device.

Abstract: The embodiments of the present disclosure disclose a signal measurement circuit and method. The signal measurement circuit include: a plurality of electrodes configured to be attached to a human body and collect physiological signals of the human body; a contact impedance detection circuit electrically connected with the plurality of electrodes, the contact impedance detection circuit being configured to measure a contact impedance between each of the plurality of electrodes and the human body; and a switching circuit configured to control a conduction state between the plurality of electrodes and the contact impedance detection circuit, such that only a portion of the plurality of electrodes are in the conduction state with the contact impedance detection circuit simultaneously.

Abstract: Embodiments of the present disclosure provide a signal acquisition system including a wearable body; a plurality of electrodes fixed on the wearable body and in contact with the skin of a user, and a processing circuit. The plurality of electrodes include a first electrode set and a second electrode set, and the first electrode set is configured to acquire a physiological signal and the second electrode set is configured to acquire a detection signal. The processing circuit configured to eliminate, based on the detection signal, motion artifacts in the physiological signal.

Abstract: Embodiments of the present disclosure disclose a signal measurement method and circuit, comprising: receiving a detection signal, the detection signal reflecting a difference between two contact impedances, each of which is a contact impedance between one of two electrodes and a human body, and the two electrodes being affixed to the human body and being configured to collect a physiological signal of the human body; and determining an indicator parameter value reflecting a quality of the physiological signal based on the detection signal. The signal measurement method and circuit provided in the present disclosure can detect the contact impedance between each of the two electrodes and the human body in real-time, and determine, based on the contact impedance, state between each of the two electrodes and the human body, and thus ensure that a physiological signal collected by the electrodes have a high quality.

Abstract: An electrocardio monitoring system and method are provided. The electrocardio monitoring system includes at least two sets of electrocardio channels and a processing circuit. Each of the at least two sets of electrocardio channels includes at least two electrodes, and each of the at least two sets of electrocardiographic channels is configured to collect signals from a human body through the at least two electrodes, respectively. The processing circuit is configured to extract the electrocardio signal from the signals collected by the at least two sets of electrocardiographic channels.

Abstract: The disclosure provides a sentence generation method based on a large language model including: obtaining a query sentence, in which the query sentence has at least one candidate description object; performing semantic recognition on the query sentence to obtain first semantic information and second semantic information corresponding to the query sentence, in which categories of the first semantic information and the second semantic information are different; inputting the at least one candidate description object and the first semantic information into the large model, to identify a target description object from the at least one candidate description object based on the large model; selecting target service data from at least one piece of service data corresponding to the target description object based on the second semantic information; and generating a reply sentence corresponding to the query sentence according to the target service data.

Abstract: One or more embodiments of the present disclosure relate to a system, a device, and a method for motion monitoring. The method includes obtaining an action signal of a user; identifying a first target interval in the action signal, the first target interval corresponding to a target action of the user; extracting, from the first target interval, a second target interval that is capable of reflecting at least one motion cycle of the target action; and evaluating the target action according to the second target interval.

Abstract: Embodiments of the present disclosure provide a method for monitoring user's running and a signal collection device. The method may include: obtaining an electromyographic (EMG) signal of a leg muscle of a user when the user is running; identifying a plurality of feature values of the EMG signal in a plurality of target time intervals; and in response to that the plurality of feature values satisfy a preset condition, generating running feedback information.

Abstract: A stair cleaning robot based on retractable rotational arms, including a cleaning robot body, wherein the cleaning robot body includes a machine body, a control system disposed at the top of the machine body, and a cleaning mechanism disposed at the bottom of the machine body; a first retractable rotational arm mechanism is disposed on the left side of the machine body, a second retractable rotational arm mechanism is disposed on the right side of the machine body, and the first retractable rotational arm mechanism and the second retractable rotational arm mechanism are disposed in a bilateral symmetry manner; the control system includes a control plate, a battery and a radar which are disposed on a control seat; a first reset sensor is disposed near the radar, and a second reset sensor is disposed near the control plate.

Abstract: One or more embodiments of the present disclosure relate to a human body posture recognition system, comprising: at least one group of ultrasonic sensors, each group of ultrasonic sensors including an ultrasonic transmitter configured to transmit ultrasonic waves and an ultrasonic receiver configured to receive the ultrasonic waves. The ultrasonic transmitter and the ultrasonic receiver are located at different parts of the body of a user, respectively; and a processor configured to, on the basis of location information of the ultrasonic transmitter and the ultrasonic receiver, information of the ultrasonic waves transmitted by the ultrasonic transmitter, and information of the ultrasonic waves received by the ultrasonic receiver, recognize the posture of the user.

Abstract: Embodiments of the present disclosure disclose a motion evaluation method, including: obtaining a motion signal of a subject, the motion signal representing a motion state of the subject; determining an evaluation criterion related to the motion signal; and evaluating the motion signal based on the evaluation criterion.

Abstract: Embodiments of the present disclosure provide a method for labeling motion data, comprising: obtaining motion data of a first subject when the first subject is in motion, the motion data representing a motion state of the first subject; obtaining image data of the first subject when the first subject is in motion; and labeling the motion data based on the image data.

Abstract: Disclosed is a physiological signal acquisition device. The physiological signal acquisition device may include a first insulating membrane, first absorbent members, and at least two electrodes. At least two first guide grooves may be arranged in a first direction on the first insulating membrane, and the at least two first guide grooves may be insulated from each other. The first absorbent members may be arranged in the at least two first guide grooves and configured to absorb moisture. The at least two electrodes may be configured to contact a skin and acquire physiological signals, wherein the at least two electrodes may cover the at least two first guide grooves and may be attached to surfaces of the first absorbent members near the skin.

Abstract: A method and system for processing gait data includes: obtaining gait data for M gait cycles of the target user's lower limbs, and determining, based on the gait data, the target lift-off moment when the target user's foot leaves the ground for each of the M gait cycles. Subsequently, gait features of the target user are determined based on at least the target lift-off moments corresponding to each of the M gait cycles. The value of the sliding factor is dynamically updated for each gait cycle to obtain the target sliding factor value. The target sliding factor value is associated with the user's motion state in that gait cycle. The lift-off time range is determined in the current gait cycle, and the target lift-off moment is determined in that gait cycle based on the target sliding factor value and the lift-off time range.

Abstract: The embodiments of the present disclosure are for a signal processing circuit. The signal processing circuit includes an analog circuit. The analog circuit is used for processing an initial signal it receives. The initial signal includes a target signal and a noise signal. The analog circuit includes a first processing circuit and a second processing circuit. The first processing circuit is used to increase a ratio of the target signal to the noise signal, and output a first processed signal. The second processing circuit is used to amplify the first processed signal. A gain multiple of the second processing circuit to the first processed signal varies with a frequency of the first processed signal. The first processing circuit includes a common mode signal suppression circuit used to suppress a common mode signal in the initial signal, a low-pass filter circuit, and a high-pass filter circuit.

Abstract: Embodiments of the present disclosure relate to devices, methods, apparatuses and computer readable storage media for controlling wireless power transfer. According to embodiments of the present disclosure, a first device (e.g., a terminal device) requests a second device (e.g., a network device) for power supply from the second device to the first device. The second device allocates a resource dedicated for a power transfer signal from the second device. The power transfer signal is used to provide power to the first device. The second device transmits configuration information indicating the dedicated resource to the first device. The first device receives the power transfer signal based on the configuration information.

Abstract: The embodiments of this application provide a method and device for optimizing neural network. The method includes: binarizing and bit-packing input data of a convolution layer along a channel direction, and obtaining compressed input data; binarizing and bit-packing respectively each convolution kernel of the convolution layer along the channel direction, and obtaining each corresponding compressed convolution kernel; dividing the compressed input data sequentially in a convolutional computation order into blocks of the compressed input data with the same size of each compressed convolution kernel, wherein the data input to one time convolutional computation form a data block; and, taking a convolutional computation on each block of the compressed input data and each compressed convolution kernel sequentially, obtaining each convolutional result data, and obtaining multiple output data of the convolution layer according to each convolutional result data.