43 results about "Voice production" patented technology

The invention discloses a voice production method based on a deep convolutional generative adversarial network. The steps comprise: (1) acquiring voice signal samples; (2) preprocessing the voice signal samples; (3) inputting the voice signal samples to a deep convolutional generative adversarial network; (4) training the input voice signals; (5) generating voice signals similar to real voice contents. A tensorflow is used as a learning framework, and a deep convolutional generative adversarial network algorithm is used to train large quantity of voice signals. A dynamic game process of a distinguishing network D and a generation network G in the deep convolutional generative adversarial network is used to finally generate a natural voice signal close to original learning contents. The method generates voice based on the deep convolutional generative adversarial network, and solves problems that an intelligent device is overly dependent on a fixed voice library to sound in a man-machine face-to-face communication process, and mode is monotonous and is lack of variations and is not natural enough.

A portable battery~powered sonic relay amplifies periodic beeping sounds from an alarm clock, smoke alarm, electronic watch or the like and analyzes the sound pulses in a logic circuit. If the sounds are periodically repeated high frequency sounds, a sound producer emits loud, low frequency sounds from a buzzer or the like, which may be heard more easily by a person with high frequency impaired hearing.

The invention discloses two making methods of handset ring sound. One is to prepare sound file in the local handset, which comprises the following steps: (1) the user plays one sound file; (2) the user intercepts random segment in the step (1) course; (3) the sound segment in the step (2) is reserved as format file of handset ring. The other is to prepare sound file from other handset color ring, which comprises the following steps: (1) the calling part user calls other handset number with color ring function, which opens local sound function to record opposite ring; (2) the calling pat user hears the color ring sound played by called part in the waiting course, which begins to record the color ring sound to generate record file; (3) the user switches the recorded color ring sound file into local handset ring format file.

The invention relates to a trash can, a control method and device of the trash can and a storage medium. A voice call signal is received, according to the voice calling signal, a voice production direction is recognized and obtained, according to the voice production direction, and a preset area map, a moving path is obtained in a planned manner, according to the moving path, a moving control instruction is sent, and the moving control instruction is used for controlling a moving device of the trash can to drive the trash can body to move to the target position. The trash can carries out direction recognizing according to the user voice calling and is moved to the position in front of a user, the user does not need to walk to the position in front of the trash can when the user needs to use the trash can, the trash can is moved to the position beside the user needing the trash can, the user can conveniently use the trash can, and the trash can use convenience is improved.

The present invention aims at the production of musical sounds by calculating motion data based on inputted image data using a simple technique without preliminarily preparing playing information or the like and by producing musical sounds based on the calculated data. A musical sound producing apparatus includes an operation part specifying means which extracts motion data indicative of motions from differentials of respective pixels corresponding to image data of a plurality of frames using image data for respective frames as an input; a musical sound producing means which produces musical sound data containing a sound source, a sound scale and a sound level in accordance with the motion data specified by the motion part specifying means; and an output means which outputs the musical sound data produced by the musical sound producing means, wherein an image database in which patterns are registered and an image matching means are provided, and a musical sound synthesizing means is provided to the musical sound producing means, in the musical sound producing means, so as to synthesize the musical sound data with other sound data, thereby producing the musical sound data.

A communications system for receiving and transmitting information signals. An electronic processor is adapted to receive information signals from at least one source, operatively connected to the electronic processor. An audible signal generator generates sounds related to the information signals. The source of information signals can be a telephone, cell phone, microphone, PDA, computer, printed document, Internet web site, e-mail or immediate message. The processor has a mechanism for generating an audible signal reminiscent of a celebrity voice, a cartoon voice, or a computer-generated sound.

The invention provides a method for rapid location and photographing through a voice instruction and a photographing system. The method comprises the steps of obtaining voice information by a robot, processing the voice information, identifying feature information, and extracting the feature information; judging whether the feature information is selfie information or not, judging that a voice production location is a photographing body if the feature information is the selfie information, identifying a location of the photographing body, otherwise, rotating a camera through a rotation device according to the feature information, thereby enabling the camera to photograph a group of reference pictures in a space of 360 degrees, or 180 degrees or 90 degrees; and photographing the picture of the photographing body. Through application of the method for rapid location and photographing through the voice instruction and the photographing system, the photographing body can be identified rapidly through the voice instruction, and the optimal photographing location can be adjusted.

The invention discloses a voice production device. The voice production device comprises a first vibration system, a second vibration system and a magnetic path system; the first vibration system includes a first vibration diaphragm as well as a voice coil arranged at one side, facing to the magnetic path system, of the first vibration diaphragm; the second vibration system includes a second vibration diaphragm arranged opposite to the first vibration diaphragm; a reinforcing portion is bonded at a middle position of the second vibration diaphragm; an extension portion extending toward a direction close to the first vibration diaphragm is arranged on the reinforcing portion; a position, corresponding to the extension portion, on the magnetic path system is provided with an avoidance portion; and the extension portion can pass through the avoidance portion and is fixed to the first vibration diaphragm. The voice production device provided by the invention is provided with two sets of vibration systems, but only a set of voice coil and magnetic path system is adopted to implement bidirectional synchronous voice production, so the structural occupied size is small and the voice production device is convenient to be widely applied to a portable terminal.

The invention discloses a mobile terminal. The mobile terminal comprises a receiving module, a conversion module, a recognition module, a modeling module and a processing module. The receiving module is used for receiving voice signals. The conversion module is used for converting the received voice signals from a time domain into a frequency domain to acquire the converted voice signals; the recognition module is used for measuring frequency response of the converted voice signals to recognize voice characteristic frequency of a voice signal initiator; the modeling module is used for building a voice production model corresponding to the voice characteristic frequency of the voice signal initiator; the processing module is used for compensating for missing components at a low-frequency part of the voice signals according to the voice production model and enhancing the strength of the voice characteristic frequency. Harmonic components of the voice characteristic frequency are added to a high-frequency part of the voice signals. The invention further discloses a voice signal processing method of the mobile terminal. By the adoption of the mobile terminal and the voice signal processing method, the processed voice signals sound more full or clearer, and user experience is improved.

The invention belongs to the technical field of information configuration software in an embedded equipment application, and is a method for performing equipment wireless network access configurationby using an acoustic wave way. The method comprises the following steps: taking a client supporting information encoding, sound manufacturing and playing as a sending end, and taking equipment end supporting sound collection and information decoding processing as a receiving end. The invention relates to the acoustic wave encoding and acoustic wave sending of the information configuration content,the sound collection of the equipment, the acoustic wave decoding and information analysis of the embedded application system on the equipment.

The invention discloses a dolby atmos processing method. The method comprises: a sound object of sound field space is obtained; a three-dimensional coordinate system is established by using a monitoring point as an original point and a three-dimensional coordinate value of the sound object is determined; the three-dimensional coordinate value of the sound object is divided into a reference block and a prediction block according to a time sequence; direct coding is carried out on the three-dimensional coordinate value of the reference block and differential coding is carried out on three-dimensional coordinate value of the prediction block; and according to the three-dimensional coordinate value before coding or after decoding, an effective acting area of the sound object is determined. According to the invention, coordination definition, moving track and acting area presentation methods during sound recording manufacturing, coding, decoding, and rendering playback processes are provided for the sound object in the three-dimensional sound field. The provided method has advantages of high coding efficiency, good sound performance, and convenient sound manufacturing.

A computer port is used to power a wireless transmission of audio signals from a sound card of a personal computer to a home entertainment system for high fidelity sound production. The computer port comprises any computer port that provides power. The computer port is used to power a FM transmitter that is connected to an audio output of a sound card of a personal computer.

The invention discloses a method for executing voice production measurement based on single-voice hole devices. The method comprises the following steps: setting voice communication parameters by virtue of a data configuration module; creating a voice progress by virtue of a data input interface, and allocating progress number to the data input interface; initializing a voice drive, invoking a speaking test staring function to create a voice stream structural body object and allocate memory to the voice stream structural body object, and filling values into the voice communication parameters; storing the voice communication parameters, and simultaneously sending signals to a business processing module; and after the signals are received by the business processing module, operating SLIC to open a media channel, and carrying out the voice production measurement. According to the method, the registering on OLT is omitted, the registering of a voice server is omitted, and single board IP is configured to connect two single-voice hole devices, so that the problem that the single-voice-hole devices cannot be subjected to a speaking test is solved, and the dependence on a specific testing environment is reduced without increasing the hardware cost.

The embodiment of the invention discloses a voice table-reservation robot. A system comprises a voice collection module, a voice recognition module, a natural language understanding module, a conversation management module, a natural language synthesis module, a voice synthesis module, a voice production module, wherein the voice collection module is used for collecting and storing voice data of auser; the voice recognition module is used for recognizing the voice data and converting the voice data into corresponding text data; the natural language understanding module is used for extractingintentions and entities in the text data; the conversation management module is used for carrying out state control, data management and context management on a conversation process; the natural language synthesis module is used for converting context output of the conversation management module into the text data; the voice synthesis module is used for converting the text data into the voice data; and the voice production module is used for playing voices. The voice table-reservation robot is capable of recognizing and understanding voice signals of a table-reservation user, extracting the intentions and entities in the voice signals and carrying out multi-round conversations and can meet the requirements of users who have clear table-reservation purposes and requires multi-round conversations.

The invention is applicable to the field of sound production, and provides a sound producing method and device. The sound producing method includes the steps of acquiring sound data; processing the tone of sound data according to a pre-established audio database; processing the volume of the sound data according to a pre-established volume threshold interval; and synthesizing the sound data using the processed tone and volume. The invention adopts the processed tone and volume to synthesize the sound data, and solves the problem that the volume and tone of the sound data cannot be automatically optimized in the prior art and therefore the improvement in the sound producing efficiency is not facilitated. The invention can automatically optimize and process the input sound magnitude and tone based on different needs so that the volume, tone, sound quality can achieve a unified standard, and has the beneficial effects of, on the one hand, improving the efficiency of sound production, and on the other hand, manually selecting the matched tone and producing a completely different sound effect.

The invention discloses a respiratory training system based on voice production. The system comprises a wearing device, a flow monitoring device, a sound collecting device and a central processing unit, wherein the wearing device is worn on a head of a patient, the flow monitoring device and the sound collecting device are arranged on the wearing device, the flow monitoring device and the sound collecting device are electrically connected or in signal connection with the central processing unit, and the central processing unit comprises an information input module, a flow monitoring module, avolume monitoring module, a training scheme module, a central processing module and a data display module. The system is advantaged in that the system requires the patient to do inspiration and pronunciation exercises at the same time, an original pure respiratory training mode is changed, the air suction amount and the sound production amount are monitored in real time, in addition, the patient can obtain corresponding visual and auditory feedback in the whole training process so that the patient adjusts the inspiration amount and the pronunciation amount in real time, the purposes of being accurate and controllable and motivating the respiratory function to the maximum degree are achieved, the training enthusiasm of the patient is further stimulated through the interesting training scheme, and then the rehabilitation training effect is improved.

The invention is applicable to the technical field of voice box equipment, and provides a vocal cavity device and a voice box. The vocal cavity device comprises a cavity, a first voice production structure and a second voice production structure, each of the first and second voice production structures is composed of a loudspeaker and a vibrating diaphragm, the loudspeakers and the vibrating diaphragms are arranged on the circumference of the cavity in axial symmetry, an included angle formed by axes of the loudspeakers of the first and second voice production structures is less than 180 degrees, the first and second voice production structures are fixed in the cavity through bolts, and an included angle formed by axes of the vibrating diaphragms of the first and second voice production structures is less than 180 degrees. A plurality of voice production structures are reasonably arranged in the cavity, so that the angle that voice sent by the loudspeakers reaches to a listener is more in line with the acoustical principle, and better acoustic effect is achieved.

Systems and methods for an enhanced ultrasound biofeedback therapy for an improved speech remediation treatment for an individual include transmitting a plurality of ultrasound (US) waves toward a tongue of the individual; receiving a plurality of reflected US waves; converting the plurality of reflected US waves into a plurality of US signals to transmit to an ultrasound machine; and generating one or more enhanced images of the tongue at least partially based on the US signals in real-time, the enhanced images including identified Regions of Interest (ROIs) along tongue sub-parts comprising the tongue root, the tongue dorsum, and the tongue blade and respective ROI points identified therein. An interactive visual story is generated and updated in real-time with a tongue-mapping trajectory of the individual on a display based on the enhanced one or more images to determine a successful or unsuccessful sound production.

The invention is to provide a dialogue robot, a dialogue system and a dialogue program with which natural dialogue with a user can be achieved. This dialogue robot i) receives answer sentence information generated by a server and including a name question for asking the name of a user if the server identifies, on the basis of facial information of the user, that the user is unregistered, and ii) receives answer sentence information generated by the server and including the name of the user if the server identifies, on the basis of the facial information, that the user is registered, and a voice production unit performs, with respect to the user, a start utterance including the name question or the name of the user according to whether the user has already been registered or not. Moreover,this dialogue robot is able to arbitrarily change the level of conversation with the user, and further may execute preprocessing such as various inquiries before changing the set level.

An electronic musical sound generator prevents a sound production sequence to be stopped from continuing to be produced even through the key is released. Even if an erroneous instruction is sent to prevent identification data from being compared, in other words, if a sound production sequence which should be stopped, continues to be produced because of failure to find the sound production sequence to be stopped, the production of the musical sound can be stopped due to the key release because a second decision block searches data in a storage block, regards a key having identification data different from the one sent as the released key, according to the sequence being produced and the key is number, and determines the sound production sequence to be stopped.

A method and device for detecting errors when practicing fluency shaping exercises, are presented. The method includes receiving a set of initial energy levels; setting a set of thresholds to their respective initial values; receiving a voice production of a user practicing a fluency shaping exercise; analyzing the received voice production to compute a set of energy levels composing the voice production; detecting based on the computed set of energy levels, the set of initial energy levels, and the set of a threshold of at least one speech-related error, wherein the detection of the at least one speech-related error is respective of the fluency shaping exercise being practiced by the user; and upon detection of the at least one speech-related error, generating a feedback indicating the at least one detected speech-related error.

A method and device for enabling remote speech disorder therapy are presented. The method includes setting a first device with at least one exercise to be performed during a current therapy session, wherein each exercise includes at least a difficulty parameter; receiving a voice production of a user of the first device; processing the received voice production to evaluate a correct execution of the voice production respective of the at least one difficulty parameter; generating a feedback based on the analysis; and outputting the generated feedback to the first device.

The invention discloses a living room multimedia computer case. The computer case comprises a case body; low-pitched voice production units symmetric to each other are fixedly connected to two sides of the outer wall of the case body through bolts; high-pitched voice production units and medium-pitched voice production units, which are symmetric mutually, are fixedly connected to the two sides of the outer wall of the case body and the upper and lower parts of the inner sides of the low-pitched voice production units; shock-absorption anti-sliding apparatuses symmetric to each other are fixedly connected to two sides of the bottom of the case body; and a display operation panel is fixedly connected to the center of the outer wall of the case body. The invention relates to the technical field of computer hardware. According to the living room multimedia computer case, the purpose of using the computer case as a video input device and an audio and video playing device of a living room television is achieved; and the shock-absorption anti-sliding apparatuses symmetric to each other are fixedly connected to the two sides of the bottom of the case body, so that the purposes of reducing the problem of case body resonance generated when the case plays voices and ensuring the stability of placement is achieved.

The invention relates to a trash can, a control method and device of the trash can and a storage medium. A voice call signal is received, according to the voice calling signal, a voice production direction is recognized and obtained, according to the voice production direction, and a preset area map, a moving path is obtained in a planned manner, according to the moving path, a moving control instruction is sent, and the moving control instruction is used for controlling a moving device of the trash can to drive the trash can body to move to the target position. The trash can carries out direction recognizing according to the user voice calling and is moved to the position in front of a user, the user does not need to walk to the position in front of the trash can when the user needs to use the trash can, the trash can is moved to the position beside the user needing the trash can, the user can conveniently use the trash can, and the trash can use convenience is improved.

An extended cognitive loudspeaker system (CLS) including a system manager coupled to a basic service set (BSS) of a wireless network, wherein the system manager establishes a CLS network using an independent BSS (IBSS) of the wireless network. A first CLS playgroup is formed by the system manager through the CLS network IBSS, wherein the first CLS playgroup includes a first control station (CS) and a first group of sound production stations (SPSs). A second CLS playgroup is formed by the system manager through the CLS network IBBS, wherein the second CLS playgroup includes a second CS station and a second group of SPSs. The second CLS playgroup can be dissolved, and the first CLS playgroup can be modified to include the second group of SPSs. The second group of SPSs can include mobility functions to enable any required movement of the second group of SPSs.

The invention discloses a dolby atmos sound coding method. The method comprises: a sound object of sound field space is obtained; a three-dimensional coordinate system is established by using a position having the same altitude as a center of a horizontal tangent plane of the sound field space and a center of a two-ear connecting line of a mixer as an original point, and a three-dimensional coordinate value of the sound object is determined; the position track of the sound object uses a frame as a unit, each frame includes a plurality of blocks and a first block of each frame is a reference block and a subsequent block is a prediction block, and a position coordinate of an ith block of the sound object is determined to be (xi, yi, and zi); and three-dimensional coordinate values of the reference blocks are coded directly and differential coding is carried out on three-dimensional coordinate values of the prediction blocks. According to the invention, coordination definition, moving track and acting area presentation methods during sound recording manufacturing, coding, decoding, and rendering playback processes are provided. The provided method has advantages of high coding efficiency, good sound performance, and convenient sound manufacturing.

The invention discloses a concentric coordinate radiation magnetic path loudspeaker which comprises a support column, cylindrical magnetic steel, a voice coil and a voice production part, wherein a step is arranged at the top end of the support column in an annular manner; the magnetic steel is arranged on the step; a gap is formed between the support column and the magnetic steel; the magnetic steel has radiation magnetic fields after radiation magnetization; the radiation magnetic fields are uniformly distributed close to the inner wall and the outer wall of the magnetic steel; the voice coil comprises a first voice coil and a second voice coil; the first voice coil is arranged inside the gap and is close to the inner wall of the magnetic steel; the second voice coil is arranged outside the magnetic steel and is adhered to the outer wall of the magnetic steel; the centers of the first voice coil and the second voice coil are overlapped; the voice production part is connected with the first voice coil and the second voice coil respectively. By adopting the concentric coordinate radiation magnetic path loudspeaker disclosed by the invention, the number of components for forming magnetic fields in a conventional loudspeaker is reduced, and the manufacturing cost of the loudspeaker is reduced.

The invention discloses a loudspeaker which comprises a front cover, a basin stand, a rear cover, a magnet block, a voice diaphragm and voice coil arranged on the inner side of the voice diaphragm. A voice producing opening is formed in the middle of the front cover, the voice diaphragm comprises a main body part in the middle and a linkage part surrounding the main body part, and the linkage part sinks corresponding to the direction, facing the rear cover, of the main body part. The linkage part on the outer periphery of the voice diaphragm is arranged to be a concave structure, a buffer gap between the front cover and the voice diaphragm is enlarged under the premise that the loudspeaker is integrally unchanged in size and the front cover is not thinned, the probability of voice distortion is effectively reduced, and voice production effect is optimized; meanwhile, a chamfering part is arranged on the inner side, opposite to the linkage part of the voice diaphragm, of the front cover, the thickness of the front cover is not increased while the rigidity of the front cover is guaranteed, and integral reliability of the loudspeaker is improved.

The invention provides a sound and picture media method for museum exhibition, and the method is characterized in that the method comprises the following steps: S1, scanning an existing cultural relic to obtain a complete cultural relic image; S2, drawing the cultural relic image to obtain a complete closed vector image line draft; S3, obtaining the color of the cultural relic according to the color of the cultural relic image and the archaeological literature; S4, splitting and synthesizing the cultural relic image to obtain a complete split image layer image; S5, carrying out skeleton structure binding on the image with the divided layers to obtain animation resources; S6, performing lighting effect production on the animation resources to obtain a dynamic image; S7, performing sound production on the dynamic image to obtain an audio file; S8, synthesizing the dynamic image and the audio file to obtain an audio-video file; S9, arranging the plurality of prism display devices in a staggered manner according to the story script; S10, performing signal output on the audio and video file, and controlling the plurality of prism display devices to perform display.