24 results about "Speech verification" patented technology

The invention relates to the technical field of voice recognition, and discloses a voice verification code generation method and system for television terminal multi-mode identity authentication. The method comprises the following steps: carrying out compensation processing on a living room environment acoustic signal to obtain a correction parameter, extracting a user voiceprint template and a spatial position feature vector to carry out multi-mode identity authentication to obtain a user identity feature code, and generating a user exclusive voice verification code audio according to the identity feature code and the voice correction parameter, and finally, the television terminal multi-mode voice verification code is associated and bound with the user identity feature code to generate a television terminal identity authentication certificate. According to the method and the device, dynamic generation of the personalized voice verification code is realized by fusing the multi-modal features such as the voiceprint and the spatial position of the user, and the security of identity authentication at the television end and the user experience are improved.

The invention discloses a forged voice voiceprint unmixing method and system, and relates to the technical field of voice processing and deep learning, and the method comprises the following steps: carrying out the feature extraction of an input forged voice based on a Transform model, and obtaining a rough feature containing the voiceprint information of a source speaker; decomposing the rough features by using a residual orthogonalization method, and recovering the voiceprint features of the source speaker; dimensionality normalization is carried out on the voiceprint features of the source speaker to obtain the voiceprint features with the fixed length; and enhancing the angle difference between the source speaker and other voiceprints by using additive angle margin loss, and outputting demixed audio data. According to the method, the influence of the voiceprint of the target speaker after voice conversion can be effectively removed, and the real voiceprint feature of the source speaker is recovered, so that the anti-counterfeiting capability of a voice verification system is improved.

The invention discloses a building construction access control method and system based on voice recognition, and relates to the technical field of voice recognition. The method comprises the steps that under the condition that a target user enters an identity verification area, a verification sequence structure is obtained, and the verification sequence structure is used for representing a priority sequence among multiple identity verification modes; under the condition that the first syn-position of the verification sequence structure is a voice recognition mode, obtaining a first to-be-verified audio input by the target user on site, a pre-stored reference audio of the target user and an environment audio of the identity verification area; based on the environment audio, performing denoising processing on the first to-be-verified audio to obtain a second to-be-verified audio; and comparing the second to-be-verified audio with the reference audio to obtain a voice verification result of the voice recognition mode, so that access of the target user is controlled based on the voice verification result. According to the invention, the efficiency and accuracy of access control can be improved.

The invention provides a VONR service capability improvement method and device, equipment, a medium and a product. The method comprises the following steps: determining VONR call ticket proportions of a plurality of different terminal types and versions based on call ticket data of a target terminal; determining VONR telephone traffic proportions of a plurality of different terminal types and versions based on the telephone traffic data of the target terminal; determining a VONR call ticket proportion threshold value based on the distribution condition of the plurality of VONR call ticket proportions, the distribution condition of the plurality of VONR call traffic proportions and the voice verification result; and on the basis of comparison of each VONR call ticket proportion and a threshold value, identifying a terminal or a version with limited VONR service capability, and generating a corresponding VONR service capability optimization strategy. According to the VONR service capability improvement method provided by the invention, the reason of the terminal or version problem can be efficiently positioned, the problem of poor VONR perception caused by the terminal or version problem is solved based on a targeted optimization strategy, and the VONR service support proportion is improved.

This invention discloses a cross-face-voice verification method and system based on trimodal fusion contrastive learning. The method includes a training phase and a testing phase. The training phase includes the following steps: S1, constructing a training sample dataset; S2, building a trimodal fusion contrastive learning model; S3, loading pre-trained parameters to accelerate model fitting and improve model training efficiency; S4, setting the parameters required for training the trimodal fusion contrastive learning model; S5, iteratively training the trimodal fusion contrastive learning model and selecting the trained trimodal fusion contrastive learning model; the testing phase specifically involves loading the parameters of the trained trimodal fusion contrastive learning model into a randomly initialized trimodal fusion contrastive learning model to complete the task of biometric feature matching. The cross-face-voice verification method based on trimodal fusion contrastive learning proposed in this invention can effectively associate face and voice data, eliminating the semantic gap between deep features.

The application discloses a kind of based on eSIM's multi-source priority automatic switching method, it is related to eSIM network switching field, including in terminal side identification user is about to initiate service whether belong to preset key service, key service includes financial payment or bank online banking or government service;Collect the historical release performance data of multiple different candidate sources in eSIM under key service, release performance data includes transaction one-time pass rate, SMS or voice verification code receiving rate and source network consistency rate;Based on the release success ability of candidate source in target service in historical release performance data is sorted;When detecting key service initiation, priority is sorted according to sorting.The application first uses the historical release performance data of key service as the core basis of network switching and priority sorting, so as to solve the problem that the traditional only depends on signal strength or time delay causes verification code receiving failure, identity discontinuity, transaction is rejected by background in key service.

The invention provides a verification code recognition method, system and device and a storage medium, and the method comprises the steps: obtaining a verification code type based on a verification code recognition request, and recognizing a layout parameter of a to-be-recognized region in an image verification code if the verification code type comprises the image verification code; and if the layout parameter satisfies a first parameter condition, performing target object identification on the image verification code by using a vision-based image identification model to obtain a target object identification result, and if the layout parameter satisfies a second parameter condition, performing target object identification on the image verification code by using a target detection algorithm to obtain a target object identification result. And if the verification code type comprises a voice verification code, recognizing the voice verification code by using a localized voice recognition model, and then simulating a corresponding user operation as a response verification code recognition request. According to the verification code identification scheme provided by the embodiment of the invention, the appropriate verification code identification strategy can be automatically selected according to the verification code type, so that the verification code identification accuracy is improved.

This application relates to a test system, method, device, and computer-readable storage medium for intelligent household appliances. The method includes: obtaining a first device state and a test instruction statement sequence of the intelligent household appliance to be tested; if the first device state is the target device state, controlling the player to play the to-be-tested instruction statement in the test instruction statement sequence; obtaining a second device state of the intelligent household appliance to be tested, and if the second device state is a non-target device state, controlling the intelligent household appliance to be tested to adjust the device state to the target device state, and performing the step of controlling the player to play the to-be-tested instruction statement in the test instruction statement sequence. In the embodiment of this application, during the voice verification test of the intelligent household appliance, by adjusting the device state of the intelligent household appliance to be tested to the target device state before each play of the to-be-tested instruction statement, the stability of the noise environment is achieved, thereby ensuring the effectiveness of the test result.

The present invention discloses a voice verification method and device based on artificial intelligence, a storage medium and an electronic device. The method comprises: obtaining a target voice generated by a target object reading a target digital string; inputting the target voice into an acoustic model to obtain multiple recognition results of the target voice and a first probability of each recognition result; calculating the second probability of each recognition result in the multiple recognition results; determining the target recognition result according to the first probability and the second probability, wherein the target recognition result is a recognition result whose second probability is less than a predetermined threshold and whose first probability is the largest; when the target recognition result is the same as the target digital string, sending a first prompt message, wherein the first prompt message is used to prompt the target object to pass the verification corresponding to the target digital string. The present invention solves the technical problem of low accuracy of voice verification.

Described herein are one or more computing devices determining, based on voice verification models, that a voice audio sample of a calling party engaging in a communication session includes characteristics of a deepfake-generated voice. In response to determining that the voice audio sample includes characteristics of a deepfake-generated voice, the one or more computing devices alert the called party about the deepfake-generated voice.

A method on a mobile device for a wireless network is described. An audio input is monitored for a trigger phrase spoken by a user of the mobile device. A command phrase spoken by the user after the trigger phrase is buffered. The command phrase corresponds to a call command and a call parameter. A set of target contacts associated with the mobile device is selected based on respective voice validation scores and respective contact confidence scores. The respective voice validation scores are based on the call parameter. The respective contact confidence scores are based on a user context associated with the user. A call to a priority contact of the set of target contacts is automatically placed if the voice validation score of the priority contact meets a validation threshold and the contact confidence score of the priority contact meets a confidence threshold.

The application relates to the technical field of voice recognition, and discloses a voice verification code generation method and system for television end multi-modal identity authentication. The method comprises the following steps: obtaining a correction parameter by performing compensation processing on a living room environment acoustic signal, extracting a user voiceprint template and a spatial position feature vector to perform multi-modal identity authentication to obtain a user identity feature code, generating a user-specific voice verification code audio according to the identity feature code and the acoustic correction parameter, performing spectrum coding processing to obtain a television end multi-modal voice verification code, and finally associating and binding the television end multi-modal voice verification code with the user identity feature code to generate a television end identity authentication credential. The application realizes dynamic generation of a personalized voice verification code by fusing multi-modal features such as a user voiceprint and a spatial position, and improves the security and user experience of television end identity authentication.

Techniques for detecting a fraudulent attempt by an adversarial user to voice verify as a user are presented. An authenticator component can determine characteristics of voice information received in connection with a user account based on analysis of the voice information. In response to determining the characteristics sufficiently match characteristics of a voice print associated with the user account, authenticator component can determine a similarity score based on comparing the characteristics of the voice information and other characteristics of a set of previously stored voice prints associated with the user account. Authenticator component can determine whether the similarity score is higher than a threshold similarity score to indicate whether the voice information is a replay of a recording or a deep fake emulation of the voice of the user. Above the threshold can indicate the voice information is fraudulent, and below the threshold can indicate the voice information is valid.

A method comprises receiving a first call from a user device operated by an emergency personnel, receiving a personal identification number associated with the emergency personnel from the user device, validating the personal identification number by verifying that the personal identification number is stored at a data store in association with an active account, wherein the active account is associated with the emergency personnel, transmitting a first prompt for voice verification registration to the user device, receiving a response from the user device to perform the voice verification registration in response to the first prompt, and either performing voice verification with the user using the user device or transmitting a prompt for a requested destination to the user device based on the response.

This application discloses a voiceprint recognition method, a graphical interface, and an electronic device. In the voiceprint recognition method, a voiceprint model is preset in the electronic device, and then the electronic device trains and updates the preset voiceprint model based on a voiceprint feature extracted from a voice of a registered user to obtain an exclusive voiceprint model belonging to the registered user. Finally, the electronic device uses the exclusive voiceprint model to generate a registered user representation based on the voiceprint feature of the voice of the registered user, and uses the registered user representation as a reference standard to realize voiceprint recognition on a voice of a speaker. Since the exclusive voiceprint model is trained based on voiceprint features of personal voices of the registered user, the registered user representation generated can accurately express voiceprint features of the user, thereby improving accuracy of voiceprint recognition.

The invention relates to a method and device for preventing a terminal serial number based on voice verification, and belongs to the technical field of terminal serial number prevention. The method comprises the steps that equipment is powered on to start network connection successfully, a random code is generated, first key information of the equipment is encrypted, and the information is sent to a platform for login authentication; the platform returns the voice code encrypted by the multi-bit random code and the random code to the equipment; bidirectional verification of the equipment and the platform is realized; encrypting second key information of the equipment, sending the information to the platform, and recording the second key information by the platform; the platform encrypts the equipment UID, the equipment password and the equipment CTEI code to the equipment, and the equipment performs login registration and is allowed to be online; and the equipment circularly broadcasts the voice codes, meanwhile, the platform is inquired whether a client carries out binding of corresponding voice code information or not through polling, when the client carries out binding matching, the platform returns equipment binding matching success, transmits a mobile phone number and a communication password, the platform records and stores equipment information, and equipment login is carried out through the password in subsequent equipment login.

The invention discloses a cross-channel invariant voiceprint feature extraction method and system based on meta learning. Relates to the technical field of speech recognition. The method comprises the following steps: step 1, acquiring multi-channel voice data; 2, constructing a meta-voice embedded network, and inputting the multi-channel voice data to train the meta-voice embedded network; step 3, optimizing the trained meta-voice embedding network, generating a training task by simulating a cross-channel scene, adjusting the meta-voice embedding network in combination with a meta optimization strategy and global distribution optimization, and generating a voice embedding space with an unchanged channel; and 4, for unknown channel voice data, voiceprint features are extracted by using the trained element voice embedding network, robust voice voiceprint embedding is generated, and cross-channel voice verification is carried out. The invention aims at improving the robustness of a voice verification system to channel mismatch and remarkably improving the voice verification accuracy under an unknown channel.

The application discloses a kind of based on voice recognition's building construction access control method and system, it is related to voice recognition technical field.The method includes: in the case where target user enters identity verification area, verification order structure is obtained, verification order structure is used to show the priority order between multiple identity verification methods;In the case where the first order of verification order structure is voice recognition method, the first audio to be verified that target user inputs on site, the reference audio of target user and the environmental audio of identity verification area are obtained in advance;Based on environmental audio, the first audio to be verified is de-noised, and the second audio to be verified is obtained;Second audio to be verified and reference audio are compared, and the voice verification result of voice recognition method is obtained, to make the access of target user based on voice verification result control.This application can improve the efficiency and accuracy of access control.

This invention relates to the field of speech verification technology, and more particularly to a speech detection method based on logarithmic graph Fourier transform feature extraction, comprising the following steps: constructing a translation operator for the speech graph, using an exponential function to describe the decay of dependencies between speech samples, generating the Laplacian matrix of the graph, and representing the speech signal as an undirected graph to capture intra-frame and inter-frame structural relationships; mapping the sample values ​​of the speech signal to graph node signals, transforming the speech signal from the time domain to the graph frequency domain, extracting frequency domain features, and forming an enhanced feature representation by synchronously merging intra-frame and inter-frame oscillation analysis and combining time domain features; generating a detection score to determine whether the speech signal is a playback attack or belongs to normal speech. This invention, by introducing logarithmic graph Fourier transform and graph signal processing methods, effectively solves the limitations of existing technologies in playback speech detection, significantly improving the comprehensiveness of feature extraction, discriminative ability, and performance of the detection system.

The application discloses a smart home voice verification method and device, equipment and medium, comprising: in response to a voice instruction initiated by a user, collecting environment context information associated with the voice instruction; based on the voice instruction and the environment context information, evaluating the security level corresponding to the voice instruction; when the security level meets the preset low-risk condition, performing basic voiceprint verification on the voice instruction, and if it matches, determining that the verification is passed; when the security level meets the preset medium-risk condition, performing dynamic voiceprint verification and context response verification on the voice instruction, and only when the dynamic voiceprint verification and the context response verification are passed, determining that the verification is passed; when the security level meets the preset high-risk condition, refusing to perform voiceprint matching operation on the voice instruction, and directly determining that the verification fails. The application can improve the security protection of voice instruction identity verification, while taking into account the use convenience of smart home voice control.

The invention provides an after-sales processing method and device of air processing equipment and control equipment. According to the method, before actual after-sales personnel provide after-sales service for target air treatment equipment, corresponding voice verification information can be automatically obtained according to voice data of the actual after-sales personnel; according to the voiceprint information of the distributed after-sales personnel of the target air treatment equipment, voiceprint verification is carried out on the voice verification information, whether the actual after-sales personnel meet operation requirements or not is determined, and a corresponding voiceprint verification result is obtained; and corresponding processing can be carried out according to the voiceprint verification result. Therefore, an effective after-sales personnel verification mechanism can be provided, the identity verification of the after-sales personnel can be quickly and accurately carried out through a voiceprint verification mode, the after-sales service quality and the user experience are ensured, and the personal and property safety of the user is guaranteed.

The invention relates to the technical field of voice verification codes, in particular to a target voice confrontation verification code generation method based on a generative adversarial network, and the method comprises the steps: obtaining a large amount of voice data, enabling the voice data to serve as original clean voice data after standardization preprocessing, and marking a transcription information label; adding disturbance guided by voice activity detection to the original clean voice data to obtain voice data with targeted disturbance preliminarily; adding a room impulse response signal to the voice data with the preliminary targeted disturbance to obtain adversarial voice data with a reverberation effect, and constructing a training sample set according to the adversarial voice data; constructing a generative adversarial network framework composed of a generator and a discriminator; inputting training samples in the training sample set into the generator, and performing confrontation training on the generator and the discriminator until convergence to obtain a trained generator; and inputting the original voice verification code and the confrontation target text into the trained generator to obtain a target voice confrontation verification code.

Communication system and related methods, in particular a method of operating a communication system is disclosed. The method comprises obtaining audio data representative of one or more voices, the audio data including first audio data of a first voice; obtaining first voice data based on the first audio data; wherein obtaining first voice data comprises applying a voice model on the first audio data; wherein the first voice data includes first speaker metric data; outputting a first voice representation indicative of the first voice data; obtaining first voice validation data, based on the first voice representation, from a first validator; obtaining second voice validation data, based on the first voice representation, from a second validator; determining an agreement metric based on the first voice validation data and the second voice validation data; determining a first validation score based on the agreement metric; and outputting the first validation score.