234 results about "Word recognition" patented technology

Automatic recognition of non-verbal messages in speech, and in particular to detection or analysis of prosodic multilayered analysis of intonation units such as prosodic unit prototypes and their multi-labeled variations, may form a hierarchical classification for the analysis of non¬ verbal information or cues in speech. A speech captured by a microphone is fed to a weakly- supervised deep learning acoustic model for speech recognition and transcription, that may be based on encoder-decoder Transformer architecture, such as Whisper by OpenAI. The model is trained to output multiple words form the text in the captured speech, to identify Intonation Units (IUs) that include one or more words, and associate non-verbal labels to each of the IUs. The labels may indicate a prototype, a discourse function (such as a conversation action), an emotion, an emphasis, or an attitude, as well as a genre of a part of, or whole of, the entire captured speech.

The invention relates to the technical field of voice decoding, in particular to a voice command word recognition post-processing method, system and device and a storage medium, and the method comprises the steps: obtaining a phoneme probability matrix and a command word path score outputted by an acoustic model; preliminarily judging whether misrecognition, mixed recognition or out-of-set word recognition exists or not based on the phoneme probability matrix and the command word path score; if the preliminary judgment result shows that misrecognition, mixed recognition or out-of-set word recognition exists, secondary confirmation is carried out based on misrecognition, mixed recognition or out-of-set word recognition, and a final recognition result is output based on a confirmation result. According to the method, the model does not need to be retrained, three types of recognition problems can be efficiently processed by utilizing a post-processing mechanism, and the recognition sensitivity and the error recognition rate can be balanced in a scene that end-side offline resources are limited.

The invention relates to the technical field of voice recognition, in particular to a mixed recognition processing method and device, equipment and a medium, and the method comprises the steps: recognizing command word data based on a command word recognition model, selecting command word data with an error rate lower than a preset range, generating a corresponding phoneme embedding representation, and forming a phoneme embedding dictionary; determining a mixed recognition command word corresponding to the current command word, and generating a mixed recognition list; identifying an audio embedding sequence corresponding to the current command word; obtaining a phoneme embedding sequence of each mixed recognition command word in the mixed recognition list based on a phoneme embedding dictionary; and calculating the similarity between an audio embedding sequence and each phoneme embedding sequence to obtain an identification result. The method and the device can improve recognition accuracy and reduce mixed recognition of command words.

The invention relates to the technical field of language data processing, in particular to a language data processing method and system, terminal equipment and a storage medium. The method comprises the following steps: acquiring current round dialogue data of a user; performing dialogue language text processing according to the current round dialogue data of the user to obtain current round dialogue language feature data; carrying out anonym detection on the current round dialogue language feature data, and carrying out dynamic historical anchor point screening to obtain cross-round screening anchor point list data; performing multi-round dialogue intention evolution reasoning according to cross-round screening anchor point list data to obtain an interaction adjustment intention representation strategy; and according to the interaction adjustment intention representation strategy, carrying out current round dialogue language key referring word determination so as to realize cross-round dialogue type language referring word identification. According to the method, the cross-round dialogue type language anonyms are accurately recognized through multi-round dialogue situation and multi-modal clue language data processing.

The invention provides a news manuscript sensitive word recognition method and system, electronic equipment and a storage medium, and relates to the field of sensitive word recognition. The method comprises the steps that internet vocabulary data is acquired and analyzed to obtain an initial sensitive word set, a sensitive word knowledge graph is obtained through knowledge graph construction and knowledge representation learning algorithm training, and a preliminary sensitive word recognition result is obtained through deep neural network integration of multi-level sensitive word recognition and an attention mechanism. And mapping the recognition result to a sensitive word knowledge graph, and performing semantic representation learning by using a graph neural network to obtain a semantic representation result. And performing association node reasoning based on the semantic representation result to obtain a knowledge enhanced recognition result. And finally, calculating the semantic similarity between the sensitive word and the context to obtain a similarity score, performing emotional tendency analysis to obtain an emotional score, and performing comprehensive evaluation in combination with a preset threshold to obtain a final sensitive word recognition result. Through the scheme, the comprehensiveness and accuracy of news manuscript sensitive word recognition are improved.

The invention belongs to the technical field of text processing, and particularly relates to a text processing method and system based on a large model, a terminal and a storage medium, and the method comprises the steps: preprocessing an input text, inputting the preprocessed text into a pre-established first large model, and outputting an error list or an error-free prompt; when the first large model outputs an error list, the output error list is detected, and the preprocessed text is corrected according to the detected error list; and inputting the corrected text or the text corresponding to the error-free prompt into a pre-established second large language model, and outputting a sensitive word prompt or a sensitive word-free prompt. According to the method, the sensitive word problem possibly introduced or triggered by the error correction operation is actively recognized and processed through the serial flow of error correction and sensitive word analysis. And the second model performs sensitive word analysis on the basis of the error-corrected text, so that the real-time accuracy and safety of sensitive word recognition are remarkably improved.

According to the embodiment of the invention, a voice processing method and device, equipment and a medium are provided. The method comprises the steps that wake-up word voiceprint features of wake-up word voice segments are extracted, the wake-up word voice segments are matched with wake-up words, and the wake-up word voiceprint features indicate a target speaker; dividing the target voice into at least one voice segment, wherein the target voice is collected after the wake-up word voice segment; extracting at least one voiceprint feature of the at least one voice segment; based on the similarity between the at least one voiceprint feature and the wake-up word voiceprint feature, determining whether the at least one voice segment is from a target speaker; and determining instruction word recognition of the at least one voice segment based on a determination result of whether the at least one voice segment is from the target speaker. Therefore, noise interference or voice interference of others can be reduced in voice wake-up and voice control applications.

The embodiment of the invention relates to the technical field of text processing, in particular to an event causal relationship identification method, device and equipment of a Chinese text. According to the scheme, the to-be-recognized Chinese text is obtained for word-by-word recognition and labeling, each word in the Chinese text is coded by adopting a new coding mode, the text code is generated, and then through the convolutional neural network and the bidirectional long-short-term memory network, the capture capability of the model on local features and context features in a long text is improved, and the recognition efficiency is improved. And fusing the local feature and the context feature vector to construct a full-text comprehensive feature representing the Chinese text, and finally generating a causal relationship labeling sequence according to the full-text comprehensive feature through a conditional random field to perform causal relationship prediction. The specific causal relationship identification weight is enhanced, the identification capability of the model on the causal relationship is improved, and the completeness and precision of the model on causal feature extraction are greatly improved.

The invention discloses a marketing mobile terminal dynamic behavior compliance monitoring method and system, belongs to the technical field of information transmission monitoring, and aims to solve the problems that existing monitoring is insufficient in non-text information processing, redundant in sensitive word judgment, incomplete in violation expression variant coverage and low in data reuse rate. The method comprises the steps that dynamic behavior information is captured, a composite retrieval code retrieval history library is constructed, and whether sensitive word recognition is conducted or not is judged; if not, generating a standardized word segmentation sequence; comparing the segmented words to determine matched segmented words and unmatched segmented words, calling a historical synonym set of the matched segmented words, positioning core semantics of the unmatched segmented words, screening evading expressions to generate a synonym set of the unmatched segmented words, and constructing a synonym data set; calling a matched word segmentation historical judgment result, establishing a synonym sequence calculation probability, and judging a sensitive word in combination with a significant proportion; executing a processing flow and updating the history library and the word segmentation library; according to the invention, efficient and accurate compliance monitoring is realized, the data reuse rate is improved, and the effectiveness of a monitoring closed loop is guaranteed.

The present application discloses a content security recognition method based on the integration of multiple large language models, belonging to the technical field of data processing. The method includes: receiving a query text, performing word segmentation processing on the query text to generate query text word segments, obtaining sensitive data texts matching the query text word segments in a sensitive database, splicing the query text and the sensitive data texts based on a preset prompt word template to generate a security recognition prompt word for the query text, inputting the security recognition prompt word into at least two large language models, obtaining the output results of the large language models, integrating the output results, and determining the security recognition result of the query text. The present application recalls sensitive data texts through a sensitive database, enhances the security recognition prompt words of the large language models, and improves the accuracy of sensitive word recognition through the integrated processing of multiple large language models.

The invention discloses a humanoid robot situational social behavior expression method, which comprises the following steps of: sensing user behaviors by adopting multiple sensors, constructing a multi-layer interactive situation sensing structure, and establishing association between a human-computer interaction situation and the user behaviors; self-adaptive layered isolated word recognition and voice emotion recognition based on an optimized neural network are adopted, user voice features are fully utilized, a matching relation between user emotion and robot situational social action emotion is constructed, and an emotion-fused robot situational social behavior expression decision is provided. Therefore, the multi-modal social behavior expression of the humanoid robot, which can express emotions while implementing an interaction task, is realized. According to the method, the relationship between the user behavior and the interaction situation and the relationship between the interaction situation and the social behavior of the humanoid robot are constructed, the problem that in the prior art, the emotion interaction is less when the humanoid robot executes the social task is solved, and the environment situation and the user emotion are fully considered; and a multi-situation social expression realization path of the humanoid robot is provided.

The present invention provides a debt collection device based on an intelligent mobile terminal, which belongs to the field of mobile communication technology, and specifically comprises: a speech-to-text conversion module is responsible for acquiring the conversation voice of a debt collector in real time and converting it into speech-to-text; a sensitive word recognition module is responsible for constructing a text record of the debt collector based on speech-to-text, communication records of instant messaging software, and text message records, and performing sensitive word recognition; a compliance evaluation module is responsible for evaluating the compliance of the debt collector; a communication frequency determination module is responsible for determining the communication frequency between the debt collector and the debtor; a repayment willingness determination module is responsible for obtaining the debtor's repayment willingness based on the number of positive keywords, the number of collection times, the weight of the positive keywords, and the debtor's repayment willingness; a result output module is responsible for warning the debt collector or suspending the collection process according to the compliance and determining the debtor's collection frequency according to the debtor's repayment willingness, thereby further reducing the compliance risk in the collection and the pertinence and efficiency of the collection.

The invention discloses a Manchu file word recognition method based on a vision-language model. The method comprises the steps of constructing an initial feature module, a stroke extraction module, a vision module, a language module, a modal alignment module and a fusion module; inputting the Manchu word picture into an initial feature extraction module to obtain an initial image feature, inputting the feature into a stroke extraction module and a visual module to obtain a stroke feature and a visual feature, inputting the visual feature into a language model, and outputting a language feature; the stroke features, the visual features and the language features are fused through a multi-mode Transform method, and a final result is controlled and output through a gating mechanism; according to the method, the recognition precision of Manchu characters with covering and staining and Manchu words with large background interference can be effectively improved.

The invention provides an out-of-vehicle voice control method and device, a storage medium and electronic equipment, and relates to the technical field of artificial intelligence, and the method comprises the steps: firstly obtaining audio signals collected by a microphone array on a vehicle for different areas outside the vehicle; the audio signals are preprocessed; performing sound source positioning and sound source separation according to the preprocessed audio signal to obtain a multi-channel audio signal; then performing wake-up word recognition based on the multi-channel audio signal, and determining a target channel where the wake-up word is located; and executing a corresponding vehicle voice control command according to the audio signal corresponding to the target channel. According to the method and the device, the audio signals are acquired through the microphone arrays in different areas around the vehicle, and the multichannel audio signals are obtained through preprocessing, sound source positioning and separation, so that accurate wake-up word recognition and voice control command execution are realized, outdoor complex environmental noise can be effectively shielded, and the accuracy of out-of-vehicle voice control is greatly improved.

The invention provides a lip language recognition method and system, terminal equipment and a medium, and the method comprises the steps: collecting a lip reading video through an event camera, and converting the lip reading video into an event sequence; voxelization is carried out on the event sequence, and data representation is carried out on the event sequence after voxelization in a voxel grid mode; and inputting the event sequence in the voxel grid form into a binary lip language recognition network model for lip language recognition to obtain a lip language recognition result corresponding to the lip reading video. According to the invention, the efficiency and accuracy of lip language recognition can be improved.

The invention discloses a content security identification method based on integration of multiple large language models, and belongs to the technical field of data processing. The method comprises the steps of receiving a query text, carrying out word segmentation processing on the query text, generating query text segmented words, obtaining a sensitive data text matched with the query text segmented words in a sensitive database, splicing the query text and the sensitive data text based on a preset prompt word template, generating security identification prompt words of the query text, and sending the security identification prompt words to the sensitive database. And inputting the security identification cue word into the at least two large language models, obtaining output results of the large language models, integrating the output results, and determining a security identification result of the query text. According to the method, the sensitive data text is recalled through the sensitive database, the safe recognition cue words of the large language models are enhanced, and the accuracy of sensitive word recognition is improved through integrated processing of the multiple large language models.

The invention relates to the technical field of natural language processing, can be applied to the fields of financial public opinion monitoring, compliance review, risk prevention and control and the like, and provides a multi-modal sensitive word processing method and device, a storage medium and computer equipment. Comprising the steps of obtaining to-be-published information and a pre-constructed knowledge base, wherein the information type of the to-be-published information comprises at least one of an image, a text, an audio and a video; performing sensitive word recognition on the to-be-published information based on a large language model and a sensitive word knowledge base to obtain a recognition result; when the sensitive content exists in the to-be-published information, determining replacement information corresponding to the sensitive content based on the enterprise-level knowledge base, the replacement information being used for replacing the sensitive content in the to-be-published information to obtain target published information; and when the sensitive content does not exist in the to-be-published information, determining the to-be-published information as target published information. According to the embodiment of the invention, automatic compliance review of cross-modal content is realized, the manual review cost is effectively reduced, and the accuracy and timeliness of sensitive information processing are improved.

The invention discloses a speech recognition method and device, a storage medium and equipment. The method comprises the following steps: firstly, obtaining a to-be-recognized target speech and extracting a target acoustic feature of the to-be-recognized target speech; inputting the target acoustic feature into a pre-constructed speech recognition model, and recognizing to obtain a speech hot word recognition result of the target speech; wherein the speech recognition model comprises an audio encoder, a hot word encoder and a decoder; the decoder includes a hotword hybrid expert adapter and an attention mechanism network. The end-to-end speech recognition large model is trained to generate the speech recognition model based on the adapter by using the sample speech data and the corresponding text labeling result and the sample hot word text, so that the model training efficiency is improved, the training cost is reduced, and the recognition accuracy of the model is also effectively improved, and therefore, the recognition efficiency of the speech recognition model is improved. When the voice recognition model is used for performing voice hot word recognition on the target voice, the efficiency and the accuracy of a recognition result can be effectively improved.

The invention discloses a vehicle-mounted scene hot word recognition method and related equipment, and relates to the technical field of Internet, and the method comprises the steps: obtaining voice data inputted by a user, extracting a hot word text from the voice data, and obtaining fusion features of the voice data and the hot word text; analyzing the hot word text to generate a plurality of mapping paths, and constructing a hot word language model based on the plurality of mapping paths; decoding the fusion feature based on the hot word language model to obtain a hot word decoding result; decoding the fusion feature based on the universal language model to obtain a universal decoding result; and determining a target recognition result based on the hot word decoding result and the general decoding result.

The embodiment of the invention discloses a voice wake-up method and device, a terminal and an earphone, and belongs to the technical field of voice wake-up. The method comprises the steps that a first audio sent by an earphone is acquired, the first audio is acquired by the earphone through a first microphone and sent when a first wake-up word recognition result of the first audio represents that a wake-up word is included, and the wake-up word is used for waking up a voice interaction function of the terminal; performing voiceprint verification on the first audio to obtain a voiceprint verification result; and under the condition that the voiceprint verification result represents that the first audio passes the voiceprint verification, waking up the voice interaction function. By adopting the scheme provided by the embodiment of the invention, the probability that the voice interaction function is mistakenly awakened can be reduced.

The invention provides a low-resource scene hot word enhancement method and device based on synthetic data, and the method comprises the steps: extracting a scene hot word list of a target scene corpus, and defining a generation task based on the knowledge classification of the scene hot word list and / or a target scene; generating a sample text set based on the first large language model application scene hot word list and the generation task; and synthesizing a sample audio set of the sample text set, and training the initial speech recognition model by taking the sample audio set as training data and the sample text set as a label to obtain a hot word recognition model. According to the method provided by the invention, a generation task is obtained based on knowledge classification of a scene hot word list and / or target scene definition; obtaining a sample text set based on the first large language model application scene hot word list and the generation task; and the sample audio set of the sample text set is synthesized, so that efficient, high-quality and diversified sample data generation in the target scene is realized, and the hot word recognition accuracy of the hot word recognition model in the target scene is greatly improved.

The invention relates to the technical field of voice recognition, and discloses a voice-control-based customized wake-up word configuration method and system for an in-vehicle machine screen, and the method comprises the steps: collecting the in-vehicle noise data of a target vehicle, and constructing a noise database; dividing noise records in the noise database into a plurality of noise levels; obtaining a user-defined wake-up word voice sample, and extracting a voiceprint feature vector of the user-defined wake-up word; measuring the similarity between the voiceprint feature vector of the custom wake-up word and each voiceprint feature vector in a voiceprint feature library, and matching to obtain a first voiceprint feature with the highest similarity; constructing a wake-up word detection sub-model corresponding to the noise level, collecting audio data in the vehicle in real time, performing wake-up word recognition according to the wake-up word detection sub-model and the first voiceprint feature, and triggering a vehicle machine screen to wake up; according to the invention, the accuracy and response speed of wake-up word recognition are improved.

The invention relates to the technical field of information query, in particular to a recruitment and application personnel information query system and method based on big data, and the system comprises a data acquisition module which is used for extracting resume data according to an application resume document and outputting a structured resume field; the data verification module is associated with an external database, verifies authenticity according to the resume field, and outputs a verification result; when the system is used, fuzzy detection, exaggerated word recognition, semantic similarity and abnormal probability distribution jointly participate in judgment, risks are stably recognized in different contexts, background cross validation and multi-dimensional fusion are carried out based on a big data model, system robustness can be improved, and man-made subjective judgment errors can be reduced; potential false information such as exaggerated packages and fuzzy descriptions is effectively recognized, the labor cost of'resumes are attractive but not consistent in ability 'is reduced, the system automatically completes preliminary screening credibility scoring, repeated consulting by workers is not needed, and the recruitment efficiency is improved.

The invention provides an isolated sign language word recognition method based on event camera space-time multi-scale feature fusion, and relates to the field of isolated sign language word recognition, and the method comprises the steps: collecting asynchronous event stream data of sign language actions through an event camera; generating an event frame sequence by using a logarithmic perception mechanism; optimizing a time multi-scale feature by using an adaptive attenuation factor; extracting spatial multi-scale features by using a feature pyramid network; and obtaining an isolated sign language word recognition result by using the full connection layer. According to the method, through time dynamic modeling and space multi-scale feature extraction of event stream data, the characterization capability of a complex time-space relationship is remarkably improved, the problems of insufficient modeling of complex time-space features of event camera data, low generalization capability and the like in the prior art are solved, and meanwhile, the defects of high data redundancy, low generalization capability and the like of a traditional frame-based method are overcome. Therefore, the recognition accuracy of the isolated sign language words is improved.

The invention provides a training method and device for a general command word recognition model, and belongs to the technical field of speech recognition, and the training method for the general command word recognition model comprises the steps: carrying out the distillation of an automatic speech recognition model, and obtaining a lightweight student model; on the basis of the student model, adding a LoRA model, a hot word post-processing module and a verification module, and constructing a general command word recognition model; and inputting the training audio into the general command word recognition model for training, and selecting different training strategies according to whether hot words are included or not in the training process. According to the training method, customization of the command words is realized through the LoRA model, so that the model has a better command word recognition effect and coverage, command word recognition and continuous speech recognition are compatible, the model optimization time and required command word training corpora are reduced, a hot word enhancement algorithm is fitted into a network, and the user experience is improved. The problem that the memory is continuously increased along with the number of hot words is solved, and the recognition speed is increased.

The invention relates to a short message copywriting sensitive word recognition and analysis method based on artificial intelligence, and provides a deep semantic modeling and cross-modal semantic alignment mechanism based on multi-modal data aiming at the technical problem of realizing rich media display and text content synthesis of marketing short messages under compliance requirements. And combining the pre-trained language model and the convolutional neural network to respectively extract high-dimensional semantic features of the text and the image, and optimizing semantic consistency through an attention mechanism and a contrast loss function. Furthermore, a dynamic sensitive word recognition and avoidance strategy is introduced into the system, sensitive expressions are effectively filtered, and compliance semantic equivalence content is generated through a text reconstruction model when necessary. And finally carrying out format fusion on the content and the image to generate a rich media short message which can be sent in compliance. According to the scheme, semantic accuracy and content compliance of multi-modal information fusion are improved, it is ensured that marketing short messages meet legal compliance requirements, and the marketing short messages have good user attraction and a good transmission effect.

The invention discloses a voice interaction method for intelligent glasses, and the method specifically comprises the steps: carrying out the wake-up word extraction of environment voice data through employing a convolutional neural network model based on an attention mechanism, and obtaining a wake-up word recognition result; verifying the wake-up word recognition result through a recurrent neural network model to obtain a wake-up word verification result, and correcting the wake-up word recognition result according to the wake-up word verification result to obtain a target wake-up word recognition result; and extracting voiceprint features when a user registers by adopting a voiceprint recognition technology, constructing a voiceprint recognition model according to the voiceprint features, and performing voice wake-up on the intelligent glasses through the voiceprint recognition model and the target wake-up word recognition result. According to the method, the accuracy, robustness and user experience of intelligent glasses wake-up word detection are remarkably improved, meanwhile, system resource consumption is reduced, and reliable technical support is provided for voice interaction of the intelligent glasses.

The invention relates to the technical field of speech recognition, in particular to a misrecognition processing method, device and equipment and a storage medium, and the method comprises the steps: scoring each entry in a command word list according to a CTC decoding rule, and determining a to-be-recognized command word; scoring a prefix part and a suffix part of the command word to be identified; judging whether the score difference between the prefix part and the suffix part exceeds a first preset threshold value or not; if the to-be-recognized command word does not exceed the preset threshold value, performing alignment processing on the to-be-recognized command word to obtain a score gap of each phoneme; and if the score difference is lower than a second preset threshold value, executing the command word to be recognized. The short command word recognition capability is enhanced, and the error recognition rate of the short command word is effectively reduced.

The system is an intelligent device for preventing and treating campus speech bullying and dirty speech scenes, sound is collected through a high-sensitivity directional MEMS microphone, a voiceprint recognition module extracts an MFCC coefficient and differential characteristics by using a librosa library, after Z-score standardization, a pre-stored voiceprint model is matched through a DTW algorithm with path constraint, and the MFCC coefficient and the differential characteristics are subjected to Z-score standardization. The integrated dirty speech recognition module constructs a dynamic multi-language lexicon (including dialects), and adopts an optimized KMP algorithm and semantic analysis to recognize improper speech. And when a dirty call is detected, the feedback module plays a warning sound through the self-adaptive volume loudspeaker and sends an alarm to the server. And the data recording module is used for locally storing data through the SD card and encrypting and transmitting the data through the Wi-Fi module. The one-key alarm physical key supports long-press triggering, pushes notifications containing positions and time and carries out voice confirmation. The equipment adopts a low-power-consumption design, the weight is less than or equal to 80g, ergonomic wearing modes such as a hanging wearing mode and a brooch type are provided, and endurance and comfort are both considered.

The invention discloses a multi-level sensitive word filtering method and system, and relates to the technical field of computers, and the multi-level sensitive word filtering method and system can comprehensively cover antagonism avoidance and semantic fuzzy scenes through a multi-level mechanism of fusion rule matching, expansion word recognition and semantic depth analysis, so that the sensitivity of a user is improved. Therefore, missed judgment and misjudgment are effectively reduced, the system adaptability is improved, missed judgment and misjudgment in the sensitive word filtering process are effectively reduced, and the adaptability and processing accuracy of the system to dynamic changes are improved.