66 results about "Normal voice" patented technology

The invention discloses a transformer fault detection method and device based on an automatic encoder and multi-scale feature fusion, and belongs to the technical field of power transformer fault detection.The method comprises the steps that sound signals during operation of a transformer are collected, and data frame division and normalization preprocessing are conducted on the collected sound signals; reconstructing the data of the collected sound signals by using an automatic encoder trained by normal sound data, and extracting the data coded by the automatic encoder as the implicit features of the sound signals; according to the error before and after reconstruction, sound abnormity judgment is carried out, and fault type detection is further carried out on abnormal data by using a classifier; and the classifier receives the collected sound data and the characteristic signal extracted by the automatic encoder as input, and classifies fault types. According to the method, the classifier is arranged, fault features are mined through the multi-scale features of the transformer sound signals, and the accuracy of transformer fault classification is improved.

Provided in the present invention are an abnormal-sound detection method and system based on diffusion denoising for an air conditioner outdoor unit. The method comprises: acquiring normal operation sound data of an air conditioner outdoor unit and interference sound data in an environment where air conditioner detection is performed; pre-processing the sound data to obtain respective time-frequency features; training a Gaussian mixture model on the basis of the time-frequency feature of the interference sound data, obtaining the data distribution thereof to perform random sampling, and superimposing, onto the time-frequency feature of normal sound data of an air conditioner, sound data sampled from the data distribution, so as to obtain fifth sound data and a corresponding sampling step label; training a neural network model by means of the normal operation sound data of the air conditioner outdoor unit and the fifth sound data, so as to obtain a trained anti-interference abnormal-sound detection model for an air conditioner outdoor unit; and acquiring operation sound data to be subjected to detection of the air conditioner, pre-processing same, and then inputting same into the trained abnormal-sound detection model for an air conditioner, so as to determine whether an operation sound of the air conditioner is abnormal. The present application improves the anti-interference performance of a detection algorithm and the air conditioner outdoor unit detection efficiency, thereby reducing the labor cost.

A method and an apparatus for improving voice call quality in an EPSFB (Evolved Packet System Fallback) scenario are provided, which are applied to the communication field. This method provides an optimization solution for refreshing a resource reservation state to resolve a problem of dual-terminal no sound, call drop, or call drop after no sound during a call between a caller terminal and a called terminal due to signaling disorder in the EPSFB scenario, effectively improving a success rate for refreshing a call state in the current scenario and ensuring a normal voice for a terminal in an abnormal scenario.

The present invention provides a method for detecting abnormal sound of multiple types of machines based on a diffusion model, comprising the following steps: collecting normal sound data, wherein the normal sound data contains category information, and constructing a training data set; preprocessing the normal sound data in the training data set to generate a mel-spectrogram, and generating a label corresponding to the category; training a multi-class abnormal sound detection network based on a diffusion model according to the label corresponding to the category and the mel-spectrogram; using the trained multi-class abnormal sound detection network to make mean‑GMM features, and training a Gaussian mixture model; inputting the sound data to be detected into the multi-class abnormal sound detection network and the Gaussian mixture model to obtain a detection result. The present invention designs an abnormal score algorithm based on a Gaussian mixture model, which evenly considers the continuity and change characteristics of the sound data in the time domain. A more fine-grained abnormal score is provided, which significantly improves the accuracy of detection and the adaptability to complex industrial environments.

Provided is an information processing device that performs processing related to voice input.The information processing device includes: a classification unit that classifies an uttered voice into a normal voice and a whisper on the basis of a voice feature amount; a recognition unit that recognizes a whisper classified by the classification unit; and a control unit that controls processing based on a recognition result of the recognition unit. The information processing device further includes a normal voice recognition unit that recognizes a normal voice classified by the classification unit, in which the control unit performs processing corresponding to a recognition result of a whisper by the recognition unit on a recognition result of the normal voice recognition unit.

The invention relates to an automatic switching method for Mesh talkback and 5G data network talkback, which comprises the following steps: a Bluetooth device creates a Mesh network and performs Mesh networking; one node is used as a main node, and an online group is created on the APP through the main node; all Bluetooth devices communicate with the APP through a bidirectional dynamic heartbeat data mode, and information synchronization is carried out on a virtual network formed by cloud mapping nodes; voice data among the online nodes are communicated in a Mesh mode; and when the node of the Bluetooth equipment is offline or reconnected, the node equipment is automatically switched to the data network mode through the APP, and networking communication is carried out by using the data network mode through the cloud. According to the automatic switching method of the Mesh talkback and the 5G data network talkback, the online condition of the equipment is monitored in real time, and when the equipment / node is offline, the cloud network is automatically switched to transmit voice data, so that normal voice data transmission and receiving of disconnected nodes are ensured, and the real-time performance and coherence integrity of voice data transmission are ensured.

The invention provides a multi-class machine abnormal sound detection method based on a diffusion model, and the method comprises the following steps: collecting normal sound data which comprises class information, and constructing a training data set; preprocessing normal sound data in the training data set to generate a Mel spectrogram, and generating labels corresponding to categories; training a multi-class abnormal sound detection network based on a diffusion model according to the labels corresponding to the classes and the Mel spectrogram; using the trained multi-class abnormal sound detection network to make mean-GMM features, and training a Gaussian mixture model; and inputting to-be-detected sound data into the multi-class abnormal sound detection network and the Gaussian mixture model to obtain a detection result. According to the invention, an abnormal score algorithm based on a Gaussian mixture model is designed, and the continuity and change characteristics of sound data in a time domain are uniformly considered. And an abnormal score with finer granularity is provided, and the detection accuracy and the adaptability to a complex industrial environment are remarkably improved.

The invention discloses a method and system for monitoring the abnormal state of a belt conveyor based on deep learning, and the method comprises the steps: laying sensing optical fibers along the line of a rack of the belt conveyor, and collecting the operation sound signals of a plurality of spatial point locations along the line through the sensing optical fibers by a distributed acoustic sensing DAS device; the method comprises the following steps: in a normal operation state of a belt conveyor, collecting normal sound signals of each point location, preprocessing the signals, and constructing a normal sample data set; for a normal sample of each spatial point location, training a corresponding unsupervised learning model; in real-time monitoring, sound signals collected at each point location are input into a corresponding unsupervised learning model, a reconstruction error is calculated, and the reconstruction error is compared with a dynamic threshold value based on normal sample statistical characteristics to obtain an abnormal result; and performing comprehensive judgment on an abnormal result based on statistical logic of the time window, and triggering an alarm if an abnormal proportion in the window exceeds a set value. High-accuracy abnormity early warning is realized, and system deployment is convenient.

A machine learning model is constructed by machine learning based on image data that is obtained by imaging each portion provided in a substrate processing device and sound data of sounds emitted when the portions operate. As the sound data, both the data of a normal sound that is emitted when each portion operates normally and the data of an abnormal sound that is emitted when each portion is abnormally operated are collected. When abnormality detection is performed, an eye tracking function identifies a position that is being viewed by a worker wearing smart glasses in which a machine learning model is installed. An image of a region including the identified viewed position is captured, and a sound is collected in a state in which the worker is viewing the viewed position. The collected image data and sound data are input into the machine learning model.

A health diagnostic system and a method for analyzing health of an animal. The system includes a signal receiver module arranged to receive a record of sound generated by an organ in a body of an animal when the organ performs a predetermined function for a predetermined period; a signal denoising module arranged to reduce a noise signal in the record; and a health diagnostic analyzing module arranged to analyze a health issue of the animal based on normal sounds generated by the organ and adventitious sounds in the denoised record.

The invention discloses a voice interaction application testing method and device and related equipment. The method comprises the steps of obtaining a target file stream from a preset file, wherein the preset file comprises a file stream generated after a normal voice interaction event stream of a voice interaction application is serialized; deserializing the target file stream to generate various voice type events; distributing the various voice type events to a voice interaction application; and driving the voice interaction application to execute the various voice type events to obtain a test result of the voice interaction application. According to the embodiment of the invention, the method does not involve the contact of an operation interface, is suitable for the test of a non-touchable voice interaction application, and guarantees the normal operation of the voice interaction application in the vehicle-mounted terminal.

The present invention provides a generative model-based method and system for detecting abnormal sounds from an air conditioner outdoor unit. The method comprises obtaining normal operating sound data of the air conditioner outdoor unit and interference sound data from the air conditioner detection environment; preprocessing the sound data to obtain time-frequency features; training a Gaussian mixture model based on the time-frequency features of the interference sound data to obtain its data distribution, which is then used for random sampling. The sampled sound data is then superimposed on the time-frequency features of the normal air conditioner sound data to obtain fifth sound data and corresponding sampling step labels. A neural network model is trained using the normal operating sound data and the fifth sound data to obtain a trained anti-interference abnormal sound detection model for the air conditioner outdoor unit; obtaining the air conditioner operating sound data to be tested, preprocessing it, and inputting it into the trained abnormal sound detection model to determine whether the air conditioner operating sound is abnormal. The present invention improves the anti-interference performance of the detection algorithm and the efficiency of air conditioner outdoor unit detection, thereby reducing labor costs.

The invention provides an interphone noise suppression method and a noise suppression device, which are used for judging whether a sound signal is noise or normal voice by utilizing a decibel value of an actually acquired sound signal and / or a ratio of a voice signal to a total sound signal, and switching on or switching off an information channel between an interphone and an earphone according to a judgment result. Moreover, the decibel threshold value of the normal voice signal is dynamically updated according to the environmental data, so that the accuracy of noise discrimination and the noise suppression effect are improved. The noise suppression device designed by the invention is externally hung between the interphone and the earphone, when the noise suppression device is used for suppressing noise, the interior of the original interphone does not need to be disassembled or maintained, and the noise suppression device is particularly suitable for maintenance of old equipment interphones.

The invention relates to a high-voltage circuit breaker spring state detection method and a state evaluation system. The method comprises the following steps: testing a subject; obtaining a sound sample of the subject in the test process and analyzing the sound sample, wherein the analysis result comprises normal and abnormal; determining the sound source position of the abnormal sound in the abnormal sound sample; repeating the test process and performing secondary sound collection on the sound source position to obtain a secondary sound sample; comparing the similarity and amplitude difference between the abnormal sound sample and the secondary sound sample; and repeating the test process again according to the amplitude difference value, correcting the position of the sound source, and marking the corrected position of the sound source. According to the high-voltage circuit breaker spring state detection method and the state evaluation system disclosed by the invention, the sound sample is obtained by driving the high-voltage circuit breaker spring to enter the simulated working state, and the sending position of the sound sample is traced by using a multi-position comparison mode, so that the automatic detection of the high-voltage circuit breaker spring is realized.

The application discloses a sound recognition alarm method and system, electronic equipment and medium, and relates to the technical field of sound recognition. A plurality of preset time periods are set, corresponding sound collection sensitivities are configured for the plurality of preset time periods and sound types respectively, and are marked in a sound existing model. Sound information in the environment is collected in real time according to the sound existing model, wherein the sound information comprises audio information, the time of collecting the sound and the sound decibel, and the sound information is constructed in the sound existing model. The audio information is analyzed to obtain the sound type, and an alarm mode is implemented on the interference sound corresponding to the sound decibel that does not satisfy the preset condition. The sound type library is obtained by storing the interference sound feature set and the normal sound feature set, has a powerful database for subsequent environmental sound recognition, can help to quickly identify the square dance music in subsequent actual life, and can efficiently react and perform alarm processing.

The invention discloses a propulsion motor nondestructive testing method, device and equipment based on an acoustic signal and a medium. The method comprises the following steps: acquiring a detection acoustic signal during operation of a propulsion motor through a sensor; converting the detection acoustic signal into a corresponding detection energy density curve; according to the deviation condition between the detection energy density curve and a reference curve in a normal sound library and between the detection energy density curve and a fault curve in a fault sound library, whether the propulsion motor has a fault or not is determined, and fault classification is determined when the fault exists; and updating the normal sound bank or the fault sound bank according to the detection energy density curve, thereby realizing early accurate identification of internal and external faults of the propulsion motor, and effectively reducing fault maintenance cost and accident risk.

The invention relates to the technical field of abnormal sound detection, in particular to a scraper abnormal sound recognition method and device and electronic equipment. The scrapper abnormal sound recognition method comprises the following steps: based on historical sound data of all working surfaces of a plurality of coal mines and an HTS-AT model, acquiring a universal audio recognition model divided according to a mining dimension; obtaining a target normal sound sample, a target fault sound sample and a target environment interference sound sample which are matched with the target working surface; based on the sampling height of the target working face, the target normal sound sample, the target fault sound sample and the target environment interference sound sample, performing fine tuning on the universal audio recognition model to obtain a target audio recognition model matched with the target working face; and based on the target audio recognition model, recognizing the to-be-detected sound of the target working surface, and obtaining an abnormal sound recognition result of the scraper. The abnormal sound of the scraper conveyor can be accurately recognized, real-time and accurate support is provided for equipment troubleshooting, and the coal mine production efficiency is improved.

This invention provides a voice interaction control method for smart home appliances, a smart home appliance, and a computer-readable storage medium. The voice interaction control method includes: turning on the smart home appliance; identifying the real-time status of a target user in the environment of the smart home appliance at preset intervals; when the real-time status of the target user is identified as active for a preset number of consecutive times, controlling the smart home appliance to execute a normal voice interaction strategy; when the real-time status of the target user is identified as sleep or rest, controlling the smart home appliance to execute a downgraded voice interaction strategy. This allows for dynamic and accurate perception of the user's real-time status to dynamically adapt to different strategies of the smart home appliance's voice interaction function, precisely preventing the smart home appliance's voice interaction function from interfering with the user's sleep or rest, and accurately realizing intelligent interaction of the voice interaction function during non-sleep or rest periods. This improves the accuracy and stability of the smart home appliance's voice interaction control, thereby enhancing the user experience.

The invention provides an axle box fault identification method, electronic equipment and a computer readable storage medium. The axle box fault identification method comprises the steps of obtaining vibration data and sound data generated by a to-be-identified axle box; wherein the axle box to be identified is a main roll shaft box in the multi-wire cutting machine; determining a normal vibration time domain signal proportion according to the amplitude of each vibration time domain signal in the vibration data and a first amplitude threshold; determining a normal sound time domain signal proportion according to the amplitude of each sound time domain signal in the sound data and a second amplitude threshold; and determining whether the axle box to be identified is in a fault state or not according to the normal vibration time domain signal proportion and the normal sound time domain signal proportion. According to the scheme, automatic fault recognition can be achieved according to the vibration data and the sound data generated in the operation process of the axle box, and therefore the accuracy and timeliness of fault detection are improved.

The invention discloses a health diagnosis system and a method for analyzing animal health. The health diagnosis system comprises: a signal receiving module for receiving a sound record generated by an organ when the organ in an animal body performs a predetermined function for a predetermined time period; the signal noise reduction module is used for reducing noise signals in records; and the health diagnosis and analysis module is used for analyzing the health problem of the animal based on the normal sound and the accidental sound generated by the organ in the record after noise reduction.

Regardless of the type of elevator, it can be easily retrofitted, detect operational abnormalities, and enable direct communication between the elevator occupants and the outside. [Solution] The elevator management device 200 is retrofitted to the outside of the elevator car and includes a microphone 204 for collecting ambient sounds, a speaker 205 for outputting sound, and a communication unit 302 for communicating with an external device 300. When the microphone 204 collects a predetermined sound or a sound similar thereto, or when the microphone 204 collects a sound different from the sounds that normally occur, the communication unit 302 transmits information about the collected sound and information about an abnormality determined based on the collected sound to the external device 300, and the speaker 205 outputs sound related to the predetermined information either towards the outside of the car or towards the inside of the car.

The invention discloses a sound transmission method based on floor obstruction in a virtual scene and a related device. The method comprises the following steps: acquiring floor space states of a target sounding object and a target listening object in a floor space environment; the floor space state shows that the target sound production object or the target sound listening object is located in the floor space, and the target sound production object and the target sound listening object are located on different floors; the normal sound resource and the floor sound resource in the floor space are mixed as a first blocking sound with relatively high blocking strength and are transmitted from the target sound production object to the target sound listening object, and the normal sound resource is strongly blocked by the floor; the floor space state indicates that the target sound production object and the target sound listening object are both located outside the floor space, and the target sound production object and the target sound listening object are located at different floors, filtering a second barrier sound with a normal sound resource being relatively low barrier strength through the floor barrier parameter, and spreading the second barrier sound from the target sound production object to the target sound listening object. Normal sound resources are weakly blocked by floors.

The invention discloses a wake-up processing method and device based on far-field voice, and terminal equipment, and relates to the technical field of voice recognition wake-up, and the method comprises the steps: carrying out the analysis processing of a detected wake-up audio, and calculating a wake-up confidence score; when the calculated wake-up confidence score of the wake-up audio is greater than or equal to a preset threshold value, determining that the wake-up is effective and controlling to enter a normal voice interaction process; when the calculated wake-up confidence score of the wake-up audio is lower than a preset threshold value, sending the wake-up audio to a background for voice recognition; performing second-level judgment: judging whether the character information of the voice recognition is consistent with a preset wake-up word or not; if so, entering a normal voice interaction process; and if not, false wake-up is confirmed, and the voice interaction process is controlled to be ended. According to the invention, the false awakening condition caused by environmental noise, similar pronunciation and the like can be solved, and convenience is provided for the use of a user.

The invention relates to the technical field of voice information security, and discloses a recording interference prevention method and device, and the method comprises an environment perception step, an equipment recognition and positioning step, an interference strategy decision-making step, a dynamic sequence generation step, a signal fusion construction step, a parametric array modulation step, a directional emission step, and an evaluation adjustment step. The device corresponds to the method. According to the application, different recording devices are accurately distinguished through an acoustic fingerprint identification technology, and a targeted interference strategy can be generated according to specific acoustic features of the devices; by combining an acoustic parametric array and a beam forming technology, accurate directional transmission of interference sound waves is realized, interference energy is enabled to act on target equipment in a concentrated manner, and the interference efficiency is improved; through fusion of multiple dynamic sequences and construction of composite baseband signals, the unpredictability and anti-filtering capability of interference signals are enhanced. Therefore, the high-efficiency interference performance is maintained, the influence on normal voice communication is reduced, and accurate and self-adaptive anti-recording protection is realized.

The invention relates to a vehicle fault early warning method and device, a storage medium and electronic equipment. The vehicle fault early warning method comprises the steps that sound signals generated when a train runs in a track preset interval are collected; fourier transform is carried out on the sound signals, and a frequency spectrum lexical element sequence is constructed based on double division of a spatial domain and a frequency domain of the sound signals; the method comprises the following steps: inputting a frequency spectrum lexical element sequence into a self-attention encoder and a self-attention decoder, and amplifying the difference between a normal sound time sequence and an abnormal sound time sequence by utilizing the comparison of different time sequence data to obtain sound signal decoding characteristics; based on the sound signal decoding characteristics output by the self-attention decoder and the fault analysis model, the fault probability, the fault type and the fault position of the train running fault obtained through sound signal analysis are determined, manual judgment is not needed in the whole process, and the fault recognition accuracy and the fault recognition efficiency can be improved.

This invention relates to the field of automotive volume control technology, and discloses a method, device, vehicle, and medium for adjusting automotive volume. The method includes: collecting ambient sound inside a target vehicle; classifying the ambient sound to obtain classification results, the classification results including normal sound data and noise data of at least one preset category; determining a noise level based on the noise category and / or the number of noises corresponding to the classification results; and adjusting the audio volume of the target vehicle based on the noise level. This invention determines the noise level by classifying the ambient sound inside the vehicle and adjusts the audio volume of the target vehicle based on the noise level. This not only achieves automatic adjustment of the vehicle volume, simplifies the adjustment process, and improves adjustment efficiency, but also provides users with a better audio experience, reduces the frequency of volume adjustments, and further ensures user safety while driving.

A health diagnostic system and a method for analyzing health of an animal. The system comprises a signal receiver module arranged to receive a record of sound generated by an organ in a body of an animal when the organ performs a predetermined function for a predetermined period; a signal denoising module arranged to reduce a noise signal in the record; and a health diagnostic analyzing module arranged to analyze a health issue of the animal based on normal sounds generated by the organ and adventitious sounds in the denoised record.

The utility model discloses an elevator control device which comprises an elevator control panel body and a plurality of floor keys arranged on the front surface of the elevator control panel body, and further comprises a sound hole square plate with a plurality of through holes formed in the surface, and a rectangular plate groove corresponding to the sound hole square plate is formed in the front surface of the elevator control panel body. A dismounting mechanism is arranged between the sound hole square plate and the elevator control panel main body; according to the elevator control panel, the elevator control panel in the prior art is improved and optimized, and the sound hole square plate convenient to disassemble and assemble is designed on the basis of the original structure, so that the sound hole square plate can be quickly disassembled for maintenance and cleaning through the disassembling and assembling mechanism when normal voice broadcast is influenced by dust accumulated in the through hole in the sound hole square plate subsequently, and the working efficiency is improved. And the whole elevator control panel main body does not need to be detached from the inner wall of the elevator car, so that the maintenance convenience of the whole device is greatly improved.

The present invention addresses the problem of providing an abnormal state detection system, an abnormal state detection method, an abnormal state detection computer-readable storage medium, and a program therefor, whereby it is possible to detect an abnormal state of a vehicle without preparing amazing data in advance. The abnormal state detection system includes: a voice abnormality detection unit including a model generation unit that generates, as a normal model, a model of normal voice data of a person taking a vehicle on the basis of voice data including voice of the person; detecting the voice abnormity based on the current voice data of the person and the normal model; a vehicle body abnormality detection unit that detects an abnormality in a vehicle body of the vehicle on the basis of acceleration data of the vehicle; and an abnormal state determination unit configured to determine whether the vehicle is in an abnormal state based on a detection result of the voice abnormality detection unit and a detection result of the vehicle body abnormality detection unit.

The utility model discloses an open type wearable earphone with a resonance structure, and relates to the field of earphones, and the earphone comprises an earphone shell which is provided with an accommodation cavity; the loudspeaker is arranged in the earphone shell and divides the accommodating cavity into a front cavity and a rear cavity; the earphone shell is provided with a sound leakage hole, and the sound leakage hole is communicated with the rear cavity. The resonance structure is arranged in the rear cavity, and the resonance structure and the sound leakage hole are arranged in a staggered manner; and the resonance structure is provided with a sealing cavity and an acoustic resistance adjusting hole, and the acoustic resistance adjusting hole is respectively communicated with the sealing cavity and the rear cavity. Compared with the prior art, sound energy limitation is not carried out on a normal sound production path, the resonance structure is arranged in the rear cavity, the main sound output path and the acoustic path of the sealing cavity are mutually independent, the normal sound production efficiency of the earphone can be fully reserved, the leakage sound is subjected to a damping effect in the transmission process through acoustic resonance, and the sound transmission efficiency is improved. Therefore, sound leakage is reduced on the premise that normal sound transmission is not interrupted and normal sound production quality is not reduced.