144 results about "Target expression" patented technology

Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for automatically generating datasets for a particular target expression photo effect. In one aspect, a system comprises receiving a plurality of image pairs, each comprising an original face image and an expressive face image representative of a target expression photo editing effect, generating an initial overall editing vector, wherein generating the initial overall editing vector comprises processing each image pair using a style space encoder model to generate an embedding of the original face image and an embedding of the expressive face image in an embedding space, optimizing the initial overall editing vector in accordance with one or more optimization criteria to generate an optimized overall editing vector, and applying the optimized overall editing vector to an input face image to generate a target expression face image that has the target expression photo editing effect.

The invention discloses a skin type robot facial expression control method and system based on a vision and language model. The method comprises the following steps of S1, obtaining an input image; capturing an image corresponding to the target expression through a camera installed at the eyes of the robot; S2, extracting facial key point features of the input image; s3, preliminarily acquiring a parameter instruction of a robot face steering engine; s4, a robot face steering engine is driven; according to the skin type robot facial expression driving method developed by the invention, visual and language information is analyzed in a staged manner, the real-time driving effect of the robot facial expression can be realized, and the driving of the robot facial expression is finely adjusted by utilizing a large language model; the robot can show good distinction degree in some subtle expressions, and the developed skin type robot has vivid facial expressions and can be applied to multiple interaction fields such as text travel, medical treatment and education.

The invention discloses an infrared and visible light image fusion method based on text supervised contrast learning. Comprising the steps that an infrared image, a visible light image and text description of the infrared image and the visible light image of the same scene are acquired, and multi-modal image features are extracted through linear embedding and cross attention Mama modules by adopting a preprocessing and image block division strategy; a pre-training text encoder is used for obtaining text semantic features, dimension alignment and weighted fusion of images and text features are achieved through an image-text feature alignment fusion module, and a fused image is generated. End-to-end training is carried out through image-text contrast learning loss in combination with pixel reconstruction and infrared intensity retention loss, and the semantic consistency and the thermal target retention capability of the fused image are effectively improved. The fusion process further comprises image-text matching weight calculation, feature weighting adjustment and fusion convolution operation. The method has the advantages of being high in semantic guiding capacity, excellent in fusion effect, accurate in hot target expression and the like, and is suitable for a multi-modal image processing task.

The invention relates to an audio data processing method, a neural network training method, an audio data processing device, electronic equipment and a computer readable storage medium. The method comprises the following steps: determining facial prior information of a virtual object based on reference video data; determining a target audio feature based on the target audio data; determining target expression sequence information based on the face prior information and the target audio feature; and determining target video data corresponding to the target audio data based on the target expression sequence information and a reference frame in the reference video data. According to the method and the device, the synchronization precision between the audio content and the facial component (such as the mouth) action of the virtual object can be remarkably improved, the real texture of the mouth action is enhanced, the expression difference between individuals is more accurately captured and reproduced, and the operation efficiency and the stability of the whole system are also improved while the performance of the system is improved.

The present disclosure relates to an audio data processing method, a neural network training method, an audio data processing apparatus, an electronic device, and a computer readable storage medium. The audio data processing method comprises: determining facial prior information of a virtual object on the basis of reference video data; determining a target audio feature on the basis of target audio data; determining target expression sequence information on the basis of the facial prior information and the target audio feature; and on the basis of the target expression sequence information and a reference frame in the reference video data, determining target video data corresponding to the target audio data. According to the present disclosure, the synchronization precision between audio content and a facial component (e.g., the mouth) action of a virtual object can be significantly improved, the sense of reality of the mouth action is enhanced, an expression difference between individuals is captured and reproduced more accurately, and the operation efficiency and stability of the whole system are improved while the system performance is improved.

The invention relates to the technical field of movie and television animation production, in particular to a movie and television animation character expression generation method and system combined with motion capture, and the method comprises the steps: firstly collecting a real-time motion data set of a target actor through motion capture equipment, including face and body motion data subsets; the facial action data is composed of facial key point displacement data with multiple continuous timestamps, then calling a preset expression mapping model to decouple the facial action data to obtain a basic and personalized expression feature set, then adjusting the basic expression feature set according to preset emotion parameters of an animation role, generating an emotion adaptive expression feature set, and generating a personalized expression feature set according to the emotion adaptive expression feature set; the method comprises the steps of obtaining a personalized expression feature set, fusing with the personalized expression feature set to obtain a target expression feature set, finally driving a three-dimensional face model to generate an expression animation by using the target expression feature set, and binding body action data to a three-dimensional body model to generate a complete character animation, thereby improving the authenticity and expressive force of animation character expressions.

The embodiment of the invention discloses a transaction risk identification method and device, a storage medium and a program product, and relates to the field of information security and artificial intelligence. The method comprises the following steps: acquiring a face image sequence and a first voice sequence of a user within a preset time length and a behavior data sequence in a target application; performing expression recognition on each frame of facial image in the facial image sequence to obtain a target expression type sequence; performing emotion recognition on the first voice sequence to obtain a target emotion type, and performing behavior recognition on the behavior data sequence to obtain a target behavior type; and performing risk analysis on the target expression type sequence, the target emotion type and the target behavior type to obtain a transaction risk value, and determining a transaction risk level of the transaction performed by the user through the target application within the preset duration based on the transaction risk value, so that risk transactions can be identified from multiple dimensions, and the user experience is improved. Especially for risk transactions of special crowds in special risk scenes, the accuracy of transaction risk identification is improved.

The invention provides a customer service platform based on an AI digital human, a service method, equipment and a medium, and relates to the field of intelligent customer service, and the platform comprises a front-end display layer which is used for generating a chat interface for interaction between a user and the AI digital human; the digital human customer service interaction layer is used for performing sentiment analysis on the query content to obtain a sentiment category corresponding to the query content; matching a target virtual image and a target expression feature corresponding to the emotion category from a configured comparison table of different emotion categories, different virtual images and different expression features; generating a target virtual image of the AI digital human according to the target virtual image and the target expression feature; outputting response content by using the target virtual image voice; and the business processing layer is used for matching response content corresponding to the query content from the configured customer service knowledge base after receiving the query request. According to the application, the AI digital human image can be combined to realize multi-terminal efficient interaction intelligent customer service.

The invention discloses an advertisement material intelligent optimization system and method based on knowledge extraction and knowledge fusion, and the method comprises the following steps: S1, obtaining and preprocessing a to-be-optimized advertisement material, and collecting an external knowledge source; s2, combined semantic analysis is executed, and cross-modal semantic alignment is carried out; s3, extracting entity units and corresponding semantic attributes, performing type labeling on the entity units, and establishing a semantic relation structure; s4, performing semantic fusion with an external knowledge source, calculating similarity between semantic vectors, performing entity mapping, performing normalization, and correcting conflict boundaries and attribute value divergence; s5, extracting a target expression fragment, and performing optimization according to a preset optimization rule set; and S6, verifying and detecting the optimized advertisement material, and outputting an optimization result. According to the method, deep understanding and cross-modal fusion optimization of advertisement materials can be realized, and the accuracy, attraction and propagation effect of advertisement semantic expression are improved.

The invention discloses a bionic robot facial expression synchronization method and device based on visual perception, equipment and a medium, and relates to the technical field of man-machine interaction, and the method comprises the steps: controlling a target bionic robot to execute a target expression control instruction used for achieving a preset expected robot facial expression, and capturing a robot visual facial expression of the robot; based on the robot visual facial expression and a preset expected robot facial expression, carrying out calibration on the target bionic robot so as to construct a hardware compensation parameter vector; executing a facial expression synchronization operation based on the hardware compensation parameter vector and a pre-trained target expression mapping model by asynchronously running a preset sensing thread and a preset driving thread; the preset sensing thread is used for collecting a target face image in real time and outputting a target steering engine control vector through a target expression mapping model; and the preset driving thread is used for outputting a steering engine control instruction at a preset output frequency based on a preset frame compensation algorithm, the hardware compensation parameter vector and the target steering engine control vector.

The invention discloses an expression recognition method and device, electronic equipment and a storage medium, and belongs to the technical field of artificial intelligence. The method comprises the following steps: acquiring an image pair comprising a target user face; the image pair comprises a natural expression image of the target user and a to-be-recognized image; performing fusion processing on the image pair, and inputting the image pair after fusion processing into a trained expression recognition model to obtain an expression recognition result of the to-be-recognized image generated by the expression recognition model; wherein the expression recognition model comprises an expression change extraction module and a decoder; the expression change extraction module is used for extracting target expression change features of the fused image pair; and the decoder is used for outputting an expression recognition result based on the target expression change feature. According to the embodiment of the invention, inherent differences of different individuals in emotional expression are fully considered, and the problem of identification deviation caused by neglecting individual expression characteristics in a traditional expression identification method is reduced, so that the accuracy of expression identification is improved.

The application provides a high-efficiency expression migration method using base expression space transformation, comprising a preprocessing stage and a facial expression migration stage; the preprocessing stage comprises: obtaining a source model O, the source model O being a base expression model comprising expression meshes under multiple different expressions; performing facial expression reconstruction on the source model O and a target model T to obtain expression description parameters of the two; performing similarity estimation according to the expression description parameters to obtain an expression parameter conversion matrix and a similarity matrix; the facial expression migration stage comprises: inputting a character picture into the base expression model, calculating expression description parameters of a target expression model according to the expression description parameters in the base expression model and the similarity matrix, and completing expression migration.

The present disclosure relates to an expression generation method and device, and a training method and device of an expression generation model, and relates to the technical field of computers. The method of the present disclosure comprises: acquiring feature information of each frame of image in an original video, feature information of a face key point, and classification information of an original expression; fusing the feature information of each frame of image, the feature information of the face key point, and the classification information of the original expression with preset classification information corresponding to a target expression to obtain feature information of a fused image corresponding to each frame of image; and generating a fused image corresponding to each frame of image according to the feature information of the fused image corresponding to each frame of image to obtain a target video in which a face expression formed by the fused images corresponding to all images is the target expression.

The present disclosure provides a landmark feature determination method and device, electronic equipment and storage medium. The landmark feature determination method comprises: obtaining target expression data, wherein the target expression data is used to represent the expression amount of each candidate feature in different cells; the average expression amount and expression proportion of each candidate feature in different cell types are counted; for each candidate feature, the first inter-class comparison data is determined according to the average expression amount of the candidate feature in the target cell type and the average expression amount of the candidate feature in the non-target cell type; for each candidate feature, the second inter-class comparison data is determined according to the expression proportion of the candidate feature in the target cell type and the expression proportion of the candidate feature in the non-target cell type; and the landmark feature of the target cell type is determined according to the first inter-class comparison data and the second inter-class comparison data. The embodiment of the present application can solve the problem of landmark feature specificity.

This invention relates to an automatic base expression selection method based on local hybrid optimization, belonging to the field of computer and automation technology. The method includes: acquiring a sequence of facial expression models with the same topology; segmenting each facial expression model into multiple sub-regions; fitting each sub-region of the base expression in the base expression group to the corresponding sub-region of each target expression model; obtaining a local facial expression model corresponding to each target expression model; fusing the obtained local facial expression models to obtain a complete facial expression model corresponding to each target expression model; comparing each complete facial expression model with its corresponding target expression model, identifying the target expression model with the largest error compared to the corresponding complete facial expression model, and adding this target expression model as the base expression to the base expression group; obtaining the base expression group when the rate of change of the largest error is less than 0.1.

The invention provides a manufacturing decision system-oriented modular functional architecture and a collaborative mechanism thereof, and the structural organization and reusable collaboration of a manufacturing decision process are realized by constructing a multi-layer capability system, a functional module system and a standardized behavior interface. The architecture comprises a data interface module, a decision-making target module, a disturbance response module and a fusion decision-making module, and defines a data flow, a control flow and a trigger relationship among the modules. The system completes target expression, constraint analysis, disturbance identification and collaborative decision generation through standardized access of multi-source data such as orders, productivity, materials and processes, and forms an interpretable and executable decision link. According to the invention, modularization, interpretability and intelligent cooperation of the manufacturing decision-making system are realized, and multi-scene and multi-industry manufacturing decision-making application is supported.

Embodiments of the present application provide a face expression migration method and an expression migration device, an electronic device and a storage medium, and belong to the technical field of artificial intelligence. The method comprises: obtaining an original face image of a target object and a target expression category; determining a first training set according to the target object, the first training set comprising a plurality of first face sample images; determining a second training set according to the target expression category, the second training set comprising a plurality of second face sample images; inputting the second face sample images into a preset initial recurrent probability model for pre-training to obtain a pre-training model; inputting the first face sample images and the second face sample images into the pre-training model to obtain a first face processing image, and then obtaining a face expression migration model; and inputting the original face image into the face expression migration model to convert the original face image into a target face image matching the target expression category. The embodiments of the present application can efficiently and accurately realize migration of a face expression in an image.

The invention discloses a kinematics semi-analytical modeling method and terminal for a human wrist-simulated parallel mechanism, and the method comprises the steps: constructing a global coordinate system, a first coordinate system and a second coordinate system when the wrist parallel mechanism is subjected to kinematics positive solution, and then obtaining a target D-H expression corresponding to the tail end pose of the human wrist-simulated parallel mechanism, a target D-H expression is obtained, the target D-H expression shows the relation between the tail end pose and a first joint angle and the relation between the tail end pose and a second joint angle, then a target modular length expression is obtained, the target modular length expression shows the relation between a first modular length and the first joint angle and the relation between a second modular length and the second joint angle, and finally, the target D-H expression shows the relation between the tail end pose and the second joint angle. And obtaining numerical values of a first modulus length and a second modulus length, and substituting the numerical values of the first modulus length and the second modulus length into the target D-H expression and the target modulus length expression to obtain a tail end pose. The invention provides a solution capable of completing operation on a low-computing-power platform for a parallel mechanism simulating a human wrist.

The application discloses a three-dimensional face generation method and device, a three-dimensional face reenactment method and device, wherein the generation method comprises the following steps: acquiring an expression vector of a first image containing a first identity object based on a target expression coding network model; the first image is an image with a face having an occlusion; inputting the expression vector and a second image containing a second identity object into a target face generation network model to determine a target image of migrating the expression vector of the first image to a face of the second identity object in the second image; and enabling the face region of the first identity object in the first image to be identified as the expression vector even in the case of having the occlusion and to be migrated to the face of the second identity object in the second image, so that the accuracy of face expression migration is improved, and the robustness of face expression migration in the case of having the occlusion is improved.

The invention discloses an information recommendation method, electronic equipment and a computer readable storage medium. The method comprises the following steps: in response to a received first data request, based on environmental parameters and position transformation information of a target object, controlling expression acquisition equipment to acquire the target object to view a target expression image of current recommendation information; performing feature extraction on the target expression image to obtain expression features of multiple dimensions; analyzing the expression of the target object for viewing the current recommendation information based on the expression features of the multiple dimensions to obtain an expression analysis result; in response to a received second data request sent by the target client, obtaining object behavior data of the target object for the current recommendation information; and adjusting the current recommendation information based on the expression analysis result and the object behavior data to obtain target recommendation information. According to the invention, the technical problem of poor information recommendation effect in related technologies is solved.

The application discloses a face expression recognition method and device, equipment and storage medium, relates to the artificial intelligence technical field, and includes: in the process of utilizing a target expression recognition model to perform face expression recognition on video frames of a target face video, for any video frame at any time, extracting each face key point, and generating a spatial heat map based on each face key point, and generating a spatial feature vector based on the spatial heat map and the image features of the video frame; determining the spatial feature vector difference between the spatial feature vector of the video frame at the current time and the spatial feature vector of the video frame at the previous time, and determining the first hidden state corresponding to the spatial feature vector of the video frame at the current time and the second hidden state corresponding to the spatial feature vector difference; determining the fusion features based on the first hidden state and the second hidden state, and determining the face expression recognition result of the video frame at the current time based on the fusion features. The application can improve the stability and accuracy of dynamic face expression recognition.

Embodiments of the present application provide a method and device for generating a dynamic image based on audio, an apparatus, and a storage medium, relating to the field of natural human-computer interaction. The method comprises: first obtaining a reference image and a reference audio input by a user; then, based on the reference image and a trained generation network model, determining a target head action feature and a target expression coefficient feature, and adjusting the trained generation network model based on the target head action feature and the target expression coefficient feature to obtain a target generation network model; finally, based on the reference audio, the reference image, and the target generation network model, processing a to-be-processed image to obtain a target dynamic image; wherein the to-be-processed image is the same as an image object in the reference image; in this way, a corresponding digital person can be obtained based on a single picture of a target person; in this way, video acquisition work and data cleaning work are not required, the production cost of the digital person can be reduced, and the production cycle of the digital person is shortened.

The invention discloses an object-driven model reconstruction method and device, a storage medium and electronic equipment. The method comprises the following steps: acquiring a first driving information set matched with a first object driving model; a first driving node set and a second driving node set are obtained from the first driving information set, the first driving node set is used for indicating a preset driving relation provided by the first object driving tool, and the second driving node set is used for indicating a self-defined driving relation included in the first object driving model; generating a driving code set according to the first driving node set, and generating a target expression set according to the second driving node set; and determining a second driving information set matched with a second object driving model according to the driving code set and the target expression set, wherein the second object driving model is used for driving the target virtual object according to a control instruction received in a second object driving tool. The method solves the technical problem of low efficiency of an object-driven model creation method.

The invention provides a statement processing method and device for a database, equipment, a medium and a product, the database can parse a to-be-processed statement to obtain a syntax tree, and an auxiliary parsing node is constructed for each expression node, so that the statement processing efficiency is improved by traversing the syntax tree. And according to the parameter class of the father node of the corresponding auxiliary node, the parameter type of the target expression node meeting the preset condition is converted, so that the parameter types of the constant and the binding parameter in the currently processed statement can be more accurately determined under the condition that the database can be used for processing data of at least two parameter types.

The application discloses a kind of kinematics semi-analytical modeling method and terminal of humanoid wrist parallel mechanism, method includes: when kinematics positive solution is carried out to wrist parallel mechanism, global coordinate system, first coordinate system and second coordinate system are constructed, then the end pose of the parallel mechanism of the humanoid wrist corresponding to target D H Expression, the target D H Expression shows the relationship between the end pose and first joint angle, second joint angle, then target module length expression is acquired, the target module length expression shows the relationship between first module length and second module length and the first joint angle and the second joint angle, finally, the numerical value of first module length and second module length is acquired, the numerical value of the first module length and the second module length is brought into the target D H Expression and the target module length expression, to acquire end pose.The application provides a solution for humanoid wrist parallel mechanism, which can be operated on a low-power platform.​​​

The digital human expression driving method, device, storage medium and computer device provided by this application, when driving the expression of a digital human, first obtain the target audio and its target text, and use the target emotion feature extraction model to extract features from the target audio and target text to obtain multi-modal features and their corresponding emotion categories, so as to improve the accuracy of expression driving through multi-modal information; then determine the human expression data corresponding to the emotion category and the target expression parameter prediction model. Since this model introduces a cross-time-and-space attention mechanism, the expression parameter sequence generated after the model performs expression parameter prediction on the multi-modal features, emotion categories and human expression data contains feature correlation operations at the time and space levels. Therefore, after obtaining the target human image, the target face generation model can be used to drive the expression of the target human image based on the expression parameter sequence to generate a digital human expression driving video with time and space continuity.

The application discloses a few-shot remote sensing image semantic segmentation method based on a double-branch reinforcement network. First, a backbone network is built to extract middle and high-level support features and query features from support images and query images respectively; second, a prototype reinforcement module based on feature customization and activation reconstruction is constructed, the middle-level support features and the query features are input into the module to obtain reinforced prototype features; then, a reinforcement module based on multi-level guidance is proposed, the high-level support features and the query features are input into the module to obtain a multi-level guidance map; finally, an encoder is designed to fuse the reinforced prototype features, the high-level query features and the multi-level guidance map to obtain a final segmentation result. The prototype reinforcement module designed by the application can deeply depict semantic features of target categories and improve target expression capability; meanwhile, the multi-level guidance reinforcement module proposed by the application can capture correlations between different levels of images, enhance the overall framework through unbiased knowledge and obtain good segmentation performance.

The present disclosure relates to a user expression-based information pushing method, device and electronic equipment, and belongs to the technical field of big data. The user expression-based information pushing method comprises: sending a pushing message to a target client; obtaining target user corresponding to the target client, expression information and operation data of the target user when the pushing message is displayed; identifying a target emotion type of the target user according to the expression information, and generating a first feedback quantization value of the pushing information based on the target expression type and the operation data; filtering candidate pushing information based on the first feedback quantization value to update a pushing information set corresponding to the target user; and selecting new pushing information for the target user from the pushing information set and returning to execute the step of sending the pushing message to the target client and subsequent steps. Thus, the present disclosure can adjust the pushing message in real time and continuously improve the mechanism of the pushing information by obtaining the user expression information and operation data in real time and analyzing the real-time user expression information and operation data, so that the pushing information is no longer limited to the user's behavior habits, the information content that can make the user happy can be pushed to the greatest extent, the accuracy and efficiency of sending the pushing message are improved, and the user experience is further improved.

The embodiment of the invention provides a special effect image generation method and device, electronic equipment and a storage medium, and the method comprises the steps: obtaining fusion deformation data and an initial face image, and enabling the fusion deformation data to comprise at least one deformation element forming a target expression; the initial face image comprises a first face element of which the element type is the same as that of the deformation element; the initial face image is processed based on the fusion deformation data, a corrected face image is generated, the corrected face image comprises second face elements, and the second face elements are matched with element contours of corresponding deformation elements in the fusion deformation data; and generating a special effect expression image with the target expression based on the corrected facial image and the fused deformation data. Before the deformation data and the face image are merged and fused, the initial face image is optimized into the corrected face image, so that the element contours of the second face element and the deformation element are matched, the face element is prevented from being covered, and the visual expressive force of the expression special effect is improved.

The invention discloses a robot facial expression generation control method based on a multi-particle swarm optimization algorithm, and relates to the technical field of robot control. According to the method, motor parameters corresponding to facial expressions of a robot are abstracted into particles, populations and parameters are initialized, expressions are generated through physical mapping, and an RGB camera captures and processes images; the method comprises the following steps: firstly, obtaining a Pareto solution set, then respectively calculating expression accuracy, key motion unit activation degree and motion unit coordination degree by using three objective functions to obtain particle fitness, then updating the Pareto solution set and particle information, finally judging whether an end condition is met, and carrying out loop iteration until a natural and vivid target expression is generated. Manual calibration is not needed, the automation degree is high, the universality is good, computing resources can be saved, and the generated expressions conform to the human face movement rule and are vivid and real.