323 results about "Interaction technique" patented technology

The invention provides an intelligent data query method based on a natural language, and relates to the technical field of intelligent data processing and natural language interaction.The intelligent data query method comprises the steps that firstly, enterprise original data is subjected to standard treatment, and a standardized theme database and a data directory and index definition document are constructed; key semantics are extracted based on unstructured knowledge, and a domain knowledge vector library is fused and constructed by combining document text fragments and vector representation of a mapping relation between historical questions of a user and an SQL (Structured Query Language). And after receiving a natural language question of a user, calling a large language model to identify a task type, and distinguishing knowledge questions and answers, data query and complex analysis. Executing corresponding operations according to different types: directly retrieving a vector library by knowledge questions and answers to generate answers; extracting keywords in data query and generating a query request in combination with context; and in the complex analysis, predefined workflow is judged and executed or intelligent agent processing is called, and query or analysis requirements are output.

The invention relates to the technical field of user interaction, in particular to a live broadcast bullet screen real-time feedback method and system based on interaction semantic matching, and the method comprises the steps: receiving and caching user bullet screens in real time, carrying out the density analysis and priority evaluation of bullet screen flows through the combination of filtering and resource optimization rules, dynamically adjusting the response frequency, and screening key bullet screens; a distillation processing mechanism intention unit is introduced, a user intention label in a key bullet screen is extracted, a vector is generated, historical behavior data of a sending user is obtained at the same time, the intention label, the vector and current real-time scene data of a live broadcast room are fused, and a bullet screen vector is constructed; performing similarity calculation in combination with the live broadcast content and user historical behaviors to obtain first feedback content; based on a reinforcement learning sentiment analysis strategy, combining user preference memory to generate second feedback content; and embedding the second feedback content into the current live broadcast interface through the client. According to the invention, real-time forward feedback and intelligent guidance of the live broadcast bullet screen are realized through interactive semantic matching.

The invention provides an interaction method, system and device and a medium, and relates to the technical field of man-machine interaction, and the method comprises the steps: obtaining an interaction process of a main control agent and a target user in a current round of interaction process according to the intention information and state information of the target user in the current round of interaction process; according to the interaction process, controlling an execution agent to perform response operation on current input information to obtain a response operation result of the current input information; the current input information is information input by the target user in a current round of interaction process; and outputting response information corresponding to the current input information according to the response operation result. According to the method, the interaction process and the dialogue content are flexibly adjusted according to the real-time intention and the real-time state of the user, and the fluency, naturalness and adaptability of interaction, the efficiency and real-time performance of interaction processing and the user experience feeling are remarkably improved.

The invention discloses a digital human interaction system based on a web end, relates to the technical field of digital human interaction, and aims to solve the problem of accumulated dislocation of browser end digital population animation and actual audio playback caused by multiple clocks and buffer scheduling. The system comprises a visual position cooperative control module, a session initialization module, an audio track and rendering canvas binding module, a multi-domain alignment time base cluster establishment module, construction of a time base cluster containing a system reference sub-time base and a content logic sub-time base, potential candidate anchor point generation module, an optimal anchor point selection module, synchronization error calculation, and judgment of a synchronization steady state, a fine adjustment state or a lost state. An optimal anchor point is screened to adjust the animation, and a visual effect studio dynamic maintenance module and an enhancement generation module assist in out-of-step processing and parameter optimization; through cooperation of multiple modules, accurate synchronization of audio and digital human animation is realized.

The invention discloses a digital large-screen interaction method and device based on multi-agent cooperation and a medium, and relates to the technical field of digital large-screen interaction.The method comprises the steps that postures, expressions and voice data of a user are collected in real time, the attention weight is calculated in combination with spatial position information, the emotional state change rate is monitored, and the user experience is improved in combination with personal historical browsing preferences; acquiring a main interaction user, an emotional state feature and an emotional change trend; according to the main interaction user, the emotional state features and the emotional change trend, obtaining an initial confidence coefficient of the suspected interested field through a semantic understanding model; and based on the candidate guide images, identifying a user selection intention through a multi-modal fusion processing mechanism, calculating a selection probability, triggering content activation of the main interaction area and content dynamic generation of the auxiliary information area, synchronizing state information, and starting an immersive collaborative interaction process. Through three-level collaborative service configuration, the beneficial effects of improving large screen content organization efficiency and enhancing immersive interaction experience are achieved.

The invention is suitable for the technical field of intelligent door locks, and provides a voice interaction processing method, device and system, an intelligent door lock and a cloud server, and the method comprises the steps: receiving a voice instruction sent by a user, and carrying out the local recognition, so as to obtain a local recognition result and the confidence of the local recognition result; when the local recognition result is matched with the local instruction set and the confidence is high, executing a local operation corresponding to the local recognition result; otherwise, establishing a secure communication session with the cloud server based on voiceprint biological characteristics extracted from the voice instruction; sending the voice instruction to a cloud server through the secure communication session, and receiving a cloud operation instruction and / or a dynamic response text returned by the cloud server; and executing corresponding operation according to the cloud operation instruction, and / or converting the dynamic response text into voice for broadcasting through a text-to-voice engine. The intelligent door lock voice interaction system solves the problem that the real-time performance, the reliability, the safety and the interaction intellectualization are difficult to balance in the existing intelligent door lock voice interaction technology.

The invention provides an AI role interaction method, interaction device and interaction equipment in a game platform, and relates to the technical field of game interaction.The method comprises the steps that a current game scene, game task information and game character information are obtained, and a first behavior parameter set of AI roles is generated; collecting current operation behavior data and voice interaction data of a user; performing intention recognition processing on the operation behavior data and the voice interaction data to generate a user behavior feature vector, and determining a relation link between the user behavior feature vector and at least one of game scene information, task information and character information to generate a dynamic interaction strategy; a second behavior parameter set is generated according to the dynamic interaction strategy, and multi-modal interaction content matched with the second behavior parameter set is rendered and generated; and displaying the multi-modal interaction content to the user terminal. According to the method, the interaction of the AI role can be more real and natural, and the game immersion and satisfaction of the user can be remarkably improved.

The invention relates to the technical field of large model interaction, in particular to a long-term and short-term memory processing system based on a large model and a vector database, and provides a scheme that short-term context memory and long-term stable memory are subjected to hierarchical management and joint retrieval by performing semantic analysis and structured modeling on user interaction contents. Under the condition that historical interaction content exists, a theme segment division and intention stack mechanism is introduced, ordered organization and step-by-step alignment of intention states in multiple rounds of conversations are achieved, and matched memory entries are obtained from a cache database and a long-term database based on query semantic representation. Furthermore, by performing normalization, situational rewriting and conflict suppression on memory entries, controllable reasoning prompt information is constructed and input to a large model to generate target output, so that the consistency, stability and controllability of a generated result are improved in a multi-round interaction and cross-session scene.

The invention discloses a voice dialogue system based on a large model, and relates to the technical field of voice interaction, and the system comprises an intelligent scheduling module which determines a task type according to a to-be-recognized text, and dynamically adapts to a fast and slow thinking module according to the task type; the fast and slow thinking module dynamically adapts to the calculation unit according to the task type and outputs a feedback result; the voice interaction module generates a to-be-recognized text according to the voice characteristics of the user and generates a voice instruction based on a feedback result; the context management module stores the interaction log to a corresponding storage unit according to the index key value; the interaction log comprises a task type belonging to the same data flow, an adaptive path and a feedback result; and the data centralized control module determines a parameter adjustment strategy according to the performance index factors in the interaction logs, and the parameter adjustment strategy acts on the intelligent scheduling module or / and the fast and slow thinking module to optimize corresponding parameters. The intelligent balance of quick response and deep reasoning of the dialogue system under a single architecture is realized, and the real-time performance and accuracy of interaction are remarkably improved.

The invention provides a vision-based cross-network interaction method and system, and relates to the technical field of intelligent interaction, and the method comprises the steps: obtaining a visual interaction sequence, constructing multi-modal feature representation, achieving cross-domain semantic alignment, deconstructing visual information into a hierarchical control instruction set, and transmitting the hierarchical control instruction set to a target network environment for execution after security classification. And a bidirectional mapping relation graph is constructed to realize incremental optimization. According to the method, semantic bridging between heterogeneous networks can be established, the cross-domain control precision is improved, and meanwhile safe interaction in a network isolation environment is guaranteed.

The invention relates to the technical field of intelligent voice interaction, in particular to an AI mobile phone voice interaction method oriented to automatic marketing, which comprises the following steps of: providing a comprehensive basis for personalized marketing and improving the marketing pertinence by recognizing a user emotional state and a real-time user intention in parallel and mastering the user state; an emotion state transition network is dynamically constructed and updated, key indexes are extracted, and a user emotion change rule is analyzed, so that an interaction strategy is accurately adjusted; semantic fragments are recombined by combining emotional stability and emotional infection degree indexes, adaptive dynamic marketing verbal skill is generated, the attraction and persuasion of the verbal skill are enhanced, and the marketing effect is improved; the local NPU is utilized to synthesize emotional voices and superpose features, so that natural and vivid interaction experience is provided, and recognition of a user on marketing content is enhanced; through dynamic decision-making verbal skill strategy cross-scene switching, timely adjustment is carried out according to user changes, marketing strategy stiffness is avoided, and the marketing success rate is guaranteed.

The invention relates to the technical field of intelligent interaction, and particularly provides a method, medium, equipment and product for tool calling in task-based dialogues, the method is applied to intelligent glasses, and the method comprises the following steps: obtaining multi-round dialogue contents of a user and a pre-trained large language model; wherein the pre-trained large language model contains a plurality of different application scenes; based on the multi-round dialogue content, a marking tool calls a cue word; outputting and calling a target tool in a scene corresponding to the tool calling prompt word by utilizing the pre-trained large language model; the target tool comprises a tool name, a tool parameter name and a tool parameter value; the target tool comprises application software corresponding to the multiple different application scenes and a function control tool of the application software. Some embodiments of the application can realize task type multi-round dialogue interaction, and meet the dialogue demand of the user.

The invention discloses a service control method based on voice analysis and a large language model, relates to the technical field of voice interaction, and solves the technical problems that in the prior art, an enterprise service system needs a user to manually fill in form fields item by item, so that operation is tedious and low in efficiency, and a process is interrupted due to lack of an intelligent completion mechanism due to lack of necessary parameters. According to the invention, a full-closed-loop control chain from the voice instruction to the service operation is constructed, the intention of the user voice input is analyzed through the fine-tuning large language model, and the service interface is dynamically matched, so that the system function operation is directly driven by the voice, and a two-way cooperation mechanism of voice input-system operation-result feedback is formed; and meanwhile, a dynamic parameter completion technology is adopted, the interface field needing to be filled is automatically verified in the parameter extraction stage, missing parameters are completed through context completion or voice guidance, the operation bottleneck of traditional form item-by-item filling is eliminated, missing of the field needing to be filled is avoided, and the service execution efficiency and the process robustness are remarkably improved.

The invention relates to the technical field of voice interaction, in particular to a dialogue type AI low-delay response control method and system based on semantic analysis. The method comprises the following steps: firstly, generating a driving intensity correlation factor of a current time window according to the change of CAN bus data; further monitoring the output of the NLU, analyzing the acoustic intonation raising characteristics, and obtaining an interaction suppression coefficient in combination with the driving intensity correlation factor; further acquiring a time pressure coefficient based on the mute duration after the user stops sounding; further comparing the interaction suppression coefficient and the time pressure coefficient of the current time window to obtain a response judgment value; further analyzing the change trend of the interaction inhibition coefficient, and updating an environment improvement flag bit; and finally, according to the driving intensity correlation factor, the environment improvement flag bit and the response decision value of the current time window, carrying out multi-level response decision control, and carrying out selective reset of an interaction state, thereby solving the problems of wrong truncation and response delay of voice interaction under dynamic driving.

The invention relates to the technical field of intelligent voice interaction, and discloses a voice interaction method and system based on 3D virtualization, and the method comprises the steps: receiving real-time voice of a user, carrying out the preprocessing, building a voice recognition model, and analyzing the pronunciation characteristics and definition of the user; optimizing the speech recognition model based on the pronunciation characteristics and definition of the user; generating a candidate prediction instruction set of the next step of the user by using the optimized voice recognition model in combination with a historical command of the user and a current 3D environment context; the candidate prediction instruction set comprises a plurality of candidate instructions; the priority of each candidate instruction is recognized, the candidate instruction with the highest priority serves as a prediction instruction of the user in the next step, related resources of the prediction instruction are loaded preferentially, and user requirements are quickly responded. The user demand is predicted more accurately in a complex scene, and the accuracy and the real-time response capability of voice interaction are remarkably improved.

The invention discloses a personalized interaction method and system based on multi-modal emotion perception, and relates to the technical field of intelligent interaction, and the method comprises the steps: collecting user historical interaction event data, and extracting interaction content features and user behavior features; obtaining interaction emotion data of the user in real time through a multi-mode sensor; fusing the interactive emotion multi-modal data to generate a dynamic emotion quantized value; generating an emotion correction factor; dynamically generating a personalized interaction strategy; and executing the interaction strategy, and collecting user response data to update a historical interaction event library. The method has the advantages that the key problem that a traditional interaction system is difficult to perceive the real-time emotional state of the user is effectively solved, a closed-loop optimization system of perception-decision-execution-feedback is constructed, and the user interaction satisfaction degree and the situation adaptive capacity of the whole system are continuously improved.

The invention provides a multi-modal fusion-driven agent interaction method and system, and relates to the technical field of intelligent interaction, and the method comprises the steps: extracting image, text and audio feature vectors through a deep neural network, carrying out the self-adaptive fusion through a cross-modal self-attention sub-module and an inter-modal interaction attention sub-module, and carrying out the self-adaptive fusion. And updating node features of the agent knowledge graph based on the difference degree, performing relation reasoning, and generating an interaction strategy to control the agent to execute an interaction action. According to the method, effective fusion of multi-modal information and dynamic updating of the knowledge graph are realized, and the accuracy and adaptability of agent interaction are improved.

The invention discloses an intelligent voice outbound system capable of realizing automatic adjustment, and relates to the technical field of intelligent voice interaction, the system comprises a voice acquisition and preprocessing module, a voiceprint analysis and emotion recognition module, a dynamic verbal skill generation module, a verbal skill real-time optimization module and a voice synthesis and outbound module; the communication efficiency and personalized service are improved through a'generation-optimization-feedback 'closed-loop design, the system generates scene-fitting verbal skills according to customer emotions, historical records and demands, adjusts strategies according to instant feedback of customers, reduces invalid conversations, reduces customer conflicts, shortens the communication period, and meanwhile, improves the user experience. Voiceprint features are decomposed into four quantifiable dimensions, emotion is accurately recognized in combination with a combination rule, limitation of a single index is avoided, the dynamic verbal skill generation module is based on a generative adversarial network, high-quality verbal skill can be learned, the dynamic verbal skill generation module adapts to multiple scenes, customization cost is reduced, and intelligent outbound large-scale application is promoted.

The invention discloses a column display sequence adaptive adjustment method and device, equipment, a medium and a product, and relates to the technical field of content recommendation and intelligent interaction. The method comprises the following steps: for each column on a platform, determining a current normalized individual interest score of a target user for the corresponding column according to user behavior data generated by the target user for the corresponding column in a plurality of current recent continuous historical unit time periods, and determining a current normalized individual interest score of the target user for the corresponding column according to a plurality of total station popularity indexes of the corresponding column in a current recent historical period; determining a current normalized global popularity score of the corresponding column, performing weighted calculation to obtain a corresponding final ranking score according to the determination result and a normalized business forced push / operation top preset compensation score of the corresponding column, and then sequentially ranking the columns according to a sequence of the final ranking scores from high to low to obtain a column sequence, and finally, the first K columns are selected from the column sequence and are displayed for the target user, so that the user experience and the platform activeness can be improved.

The present disclosure provides an electronic device (100) for handling an anomaly detection. The electronic device includes a data integration engine (120) obtaining a user-specific data from a plurality of sources. Further, the electronic device includes an LLM engine (125) utilizing an LLM technique and a natural language processing framework to process and analyse textual interactions included in the user-specific data. The LLM technique extracts linguistic cues and patterns from the user-specific data and creates a behavioural biometrics profile for each user from a plurality of users. Further, electronic device includes a contextual understanding engine (130) providing contextual understanding of user behaviour by considering semantics and sentiment of textual interactions and voice data included in the user-specific data using LLM technique. Further, the electronic device includes an anomaly detection engine (135) triggering anomaly detection using LLM technique by comparing the user behaviour with the collected user- specific data.

The embodiment of the invention discloses an information display method and device and a computer program product, and relates to the technical field of information interaction. The method comprises the steps that real-time state information used for representing the running state of a target object is displayed on a target control of a graphical user interface; in response to the obtained suggested state information related to the running state, target visual elements are displayed on the target control, the target visual elements at least comprise a first visual element and a second visual element, the first visual element is used for indicating the real-time state information of the target object, and the second visual element is used for indicating the real-time state information of the target object. The second visual element is used for indicating suggested state information of the target object, the display position relation of the first visual element and the second visual element on the target control is determined by a first comparison result, and the first comparison result is a real-time comparison result of the real-time state information and the suggested state information. According to the method and the device, information display can be more abundant and reasonable, so that a user is helped to make a decision efficiently and accurately.

The invention discloses a multi-modal driving system based on streaming big language model output, which belongs to the technical field of multi-modal interaction and comprises the following steps: a streaming big language model module receives input information and then generates a text unit sequence in a streaming manner; when each text unit is generated, a multi-dimensional emotion label sequence containing emotion types, emotion intensity values and semantic association degrees is synchronously output; the speech generation module maps the text unit sequence and the corresponding multi-dimensional emotion label sequence into speech synthesis parameters, and drives a streaming text-to-speech model to generate an emotional speech stream; the label generation module generates an expression label sequence and an action label sequence of the virtual image in real time; the emotion synchronous control module performs synchronous interpolation and coordination control by taking the time step of the text unit as a reference to generate a synchronous multi-mode driving signal; and the rendering and driving module drives the virtual image based on the signal to realize collaborative output of voice, mouth shape, expression and limb movement, and realizes synchronous presentation of multi-modal emotion expression.

The invention relates to the technical field of intelligent interaction, in particular to an interaction system based on emotion recognition, and the system comprises the steps: collecting multi-modal sensing data and environment context of a user; the feature matching module is used for constructing a first situation feature by utilizing each single-dimensional feature deviation degree obtained by calculating the current multi-modal sensing data of the user and a corresponding coupling feature deviation degree when the first situation is matched, and constructing a second situation feature when the first situation is not matched; constructing a second situational feature by using the deviation degree of each single-dimensional feature obtained by calculating the current multimode sensing data of the user; the emotion risk index calculation module is used for fusing the first situation feature or the second situation feature according to a pre-trained fusion model to obtain an emotion risk index; and the interaction module is used for triggering the interaction strategy of the corresponding level according to the relationship between the emotion risk index and the latest emotion risk threshold. In the emotion recognition process, the current situation type of the user is fused, and the accuracy of emotion recognition is greatly improved.

The invention discloses an intelligent interaction method and system for a meta-universe scene, and relates to the technical field of meta-universe scene interaction. The method comprises the following steps: acquiring scene interaction information and performing information analysis to obtain a voice analysis result, an action analysis result and a face analysis result; performing multi-modal sentiment analysis based on an analysis result to obtain a sentiment intention of the user; interacting with the user according to the emotional intention to obtain first interaction feedback information; acquiring a scene interaction log and performing log analysis to obtain an interaction recommendation result; interacting with the user according to the interaction recommendation result to obtain second interaction feedback information; and comprehensively analyzing the interaction feedback information and updating and optimizing the interaction process. According to the method, from two perspectives of scene interaction information and scene interaction logs, emotional intentions of users are obtained by adopting multiple technologies such as deep learning and neural networks, so that the accuracy of results, the usability and selectivity of interaction and the experience effect of the users are improved to a great extent.

The invention relates to the technical field of virtual exhibition hall interaction, and discloses a virtual exhibition hall interaction method and system. The method comprises the steps of firstly obtaining user interaction data collected by various terminals such as touch, voice and somatosensory terminals, extracting behavior characteristics and matching to obtain interaction pairs, and specifically, extracting characteristic descriptors, distributing time sequence labels and searching matched nodes to obtain the interaction pairs. And then constructing an interaction model by using the user interaction data and the interaction pairs, and performing scene encapsulation to generate a standard interaction packet, including obtaining the interaction model by taking the interaction pairs as anchor points, and performing layered scene encapsulation through two levels. In addition, topological data of the exhibition hall is obtained to generate an interaction path, and multiple terminals collect data in sequence and screen interaction pairs. The system comprises an interaction controller which can execute the operation. According to the invention, effective processing of multi-terminal interaction data and construction of the interaction model are realized, and the fluency and accuracy of interaction of the virtual exhibition hall are improved.

The invention discloses a dynamic interaction system, an interaction execution mechanism and intelligent equipment, and belongs to the technical field of dynamic interaction.The dynamic interaction system comprises the steps that user operation requirements are received, data are collected through a basic platform pressure sensor to generate a matrix, pressure stability is analyzed through a time window, data are processed only in the stable period, and a contact body contour is extracted; and synchronously judging a placement state and generating a calibration signal. The extracted contour is matched with a standard contour feature point, and a precise operation point coordinate and a control instruction are generated in combination with a geometrical relationship and elastic transformation. In the execution link, the execution unit is driven to operate the contact body, and the adjustable unit dynamically compensates the low-pressure area and the supporting area to ensure uniform stress and stability. And the interaction feedback module guides, adjusts and visualizes the state through a multi-modal interface. The scheme improves identification precision and operation stability through dynamic pressure analysis, intelligent contour verification and the like, is suitable for medical, industrial and other scenes, and provides an efficient intelligent solution.

The invention discloses a user deep interaction system and method based on a multi-source data platform, and relates to the technical field of user deep interaction, and the method comprises the steps: obtaining a historical online interaction record and a historical offline interaction record of a user in a shopping mall, evaluating the interaction tendency of the user to commodities in the shopping mall, and obtaining interaction tendency data; obtaining interaction tendency data of each user in the shopping mall, and analyzing the overall interaction tendency of each user to commodities in the shopping mall to obtain target interaction tendency data of the shopping mall; acquiring historical consumption records of the user in the shopping mall, and analyzing consumption collocation of commodities of the user in the shopping mall to obtain consumption collocation data of the user; and acquiring consumption matching data of each user in the shopping mall, analyzing consumption matching degrees among different commodities in the shopping mall, and optimizing and adjusting an interaction strategy of the shopping mall in combination with the target interaction tendency data.

The invention discloses an intelligent interaction system and method based on dynamic intention recognition, and relates to the technical field of intelligent interaction.The intelligent interaction method comprises the steps that firstly, current query of a user and a context set including historical dialogues, browsing records and other information are obtained; a parallel embedding model is introduced to extract semantic features of user query and semantic features of each piece of information in a context set, and dynamic correlation weighting and screening are performed on context features through a gating network with current query as guidance, so that self-adaptive deep fusion of key context information and user query is realized; and performing intention classification inference on the basis of the semantic representation after fusion enhancement to obtain a final intention recognition result. Therefore, the problems of insufficient utilization of the context information and serious noise interference can be solved, so that the intention understanding accuracy and robustness of the system when the system processes the reference unknown query are remarkably improved.

The invention relates to the technical field of somatosensory interaction, and discloses a holographic naked eye teaching experiment module somatosensory interaction method and system, and the system comprises a virtual environment scene construction module, a holographic naked eye scene display module, a somatosensory behavior feature recognition module, a somatosensory interaction behavior recognition module, an interaction command conversion module, and a multi-mode feedback module. A depth coordinate system is established through a somatosensory camera to perform joint point recognition on a user and obtain a user behavior depth image, user behavior feature extraction is performed based on the user behavior depth image, a tracking algorithm is utilized to perform real-time tracking on the user behavior features, and a somatosensory interaction behavior recognition result is obtained through analysis. The somatosensory interaction behavior is converted into the corresponding operation command based on the somatosensory interaction behavior recognition result, accurate recognition and quick response to the user behavior are achieved, and the accuracy and efficiency of somatosensory interaction are effectively improved.