46 results about "Emotion perception" patented technology

The invention discloses a home service robot cloud multimode dialogue method, device and system. The method includes the steps: receiving user voice information and scene image information in real time; converting the user voice information into text information, and performing word segmentation and named entity identification to determine dialogue types; preprocessing the scene image information;screening preprocessing image information and processed text information, adding the information into an incremental knowledge base, performing incremental model training when dialogue service load rate is smaller than a threshold value, and updating a dialogue service model; performing scene perception, user perception and emotion perception on the preprocessing image information and/or processed text information to obtain scene perception information, user perception information and emotion perception information; combining the processed text information with the scene perception information, the user perception information and/or the emotion perception information according to different dialogue types to generate interactive dialogue information of a user and a robot through the dialogue service model.

The invention relates to an emotion perception interactive amusement apparatus that uses the inertia sensing module to sense the user's action as the input judgment of game control, and adds the physiological sensing module to sense the physiological morphological feature of heartbeat and skin as the emotion state. The emotion perception interactive amusement apparatus can unscramble the signal sensed by the inertia sensing module and the physiological sensing module to know the real-time emotion and action, and feedback to the game platform to generate the interactive entertainment effect, and it can train the coordination of user's body and control the self emotion to provide a body and mind amusement apparatus.

A property vector representing extractable measurable properties, such as musical properties, of a file is mapped to semantic properties for the file. This is achieved by using artificial neural networks “ANNs” in which weights and biases are trained to align a distance dissimilarity measure in property space for pairwise comparative files back towards a corresponding semantic distance dissimilarity measure in semantic space for those same files. The result is that, once optimised, the ANNs can process any file, parsed with those properties, to identify other files sharing common traits reflective of emotional perception, thereby rendering a more liable and true-to-life result of similarity/dissimilarity. This contrasts with simply training a neural network to consider extractable measurable properties that, in isolation, do not provide a reliable contextual relationship into the real-world.

The invention relates to the technical field of a hotel service robot system based on cloud voice communication and emotion perception. The hotel service robot system includes a cloud voice recognition module, a semantic understanding module, a voice playing module, a facial expression recognition module, a face recognition module, a display module, an action recognition module, an emotion recognition module, a check-in/check-out module, a print material module, a payment module, a surveillance video system module, an identity card scanning module and the like. The above modules are integratedin a hotel service robot, the hotel service robot judges whether there is a customer check-in and check-out service by using a face recognition technology, and handles the service by voice communication with a customer, and meanwhile, the hotel service robot has functions of printing of related materials, payment, identity scanning and the like. The hotel service robot judges customer emotion through three aspects of information of human facial expression, voice and action to properly adjust emotional scenes for communication with the customer, and the hotel service robot can call a hotel monitoring system to perform face recognition in order to remind customers whose face images are not collected and recorded to collect and record the face images.

The invention discloses a body language-based emotion perception method adopting deep learning and UKF. The method comprises the following steps of monitoring a person entering a Kinect work area by adopting Kinect, and then calculating skeleton points of the person in real time; estimating the positions of the skeleton points by using unscented Kalman filtering, and calculating a measurement error generated by a tracking error and noises of a device; adopting a convolutional neural network method for static body actions, and adopting bidirectional long-short-term memory conditional random field analysis for dynamic body actions; and directly putting output items of characteristics obtained after action processing into a softmax classifier for performing identification, thereby identifyingeight emotions. The body language-based emotion perception has the following advantages that firstly, body languages can be more easily captured by a sensor; secondly, emotion perception-based body language noises are relatively small; thirdly, the body languages have less fraudulence; and fourthly, the capture of the body actions does not influence or disturb the actions of participants.

The invention discloses an internet-of-things emotion perception technology-based internet-of-things social application system. The system comprises a hardware acquisition module and a social APP, wherein the hardware acquisition module is used for measuring and recording heart rate and shell temperature data of users in real time and sending the data to a mobile terminal; the social APP is installed on the mobile terminal; and the social APP has a basic chat function, is internally provided with an emotion perception algorithm, and is used for receiving the data uploaded by the hardware acquisition module in real time after being correctly matched with the hardware acquisition module, analyzing emotion of a user using the current hardware, and transferring emotion perception information through contour pattern change or vibration so as to realize emotion perception and interaction. Aiming at the current situation that the current internet-of-things social application software is lack of scientific measurement and real-time emotion change of interaction users, the method fuses an internet-of-things emotion perception technology into internet social applications, so that the dead zone that the current internet social applications ignore the user emotion interaction is filled, the interaction experience in internet social contact is increased and the social contact quality is enriched.

A property vector representing extractable measurable properties, such as musical properties, of a file is mapped to semantic properties for the file. This is achieved by using artificial neural networks “ANNs” in which weights and biases are trained to align a distance dissimilarity measure in property space for pairwise comparative files back towards a corresponding semantic distance dissimilarity measure in semantic space for those same files. The result is that, once optimised, the ANNs can process any file, parsed with those properties, to identify other files sharing common traits reflective of emotional-perception, thereby rendering a more liable and true-to-life result of similarity/dissimilarity. This contrasts with simply training a neural network to consider extractable measurable properties that, in isolation, do not provide a reliable contextual relationship into the real-world.

The invention discloses a dialogue emotion recognition network model based on double knowledge interaction and multi-task learning, a construction method, electronic equipment and a storage medium, and belongs to the technical field of natural language processing. The problem that an existing Emotion Recovery in Conversion (ERC) model neglects the direct interaction between the utterance and the knowledge is solved. And when an auxiliary task weakly related to the main task is used, limited emotion information can only be provided for the ERC task. According to the method, common knowledge in a large-scale knowledge graph is utilized to enhance word level representation. A self-matching module is used to integrate knowledge representations and utterance representations, allowing complex interaction between the two representations. And a phrase-level sentiment polarity intensity prediction task is taken as an auxiliary task. A label of the auxiliary task comes from an emotion polarity intensity value of an emotion dictionary and is obviously highly related to an ERC task, and direct guidance information is provided for emotion perception of a target utterance.

The invention discloses a social network interpersonal relationship analysis method. The method comprises the steps of firstly collecting data of microblog users; according to basic information and interactive contents of the microblog users, obtaining a primary relationship network of the microblog users; based on the primary relationship network, generating a directed connected graph; analyzing the interactive contents to obtain an interactive frequency and interactive intimacy; according to the interactive frequency and the interactive intimacy, obtaining a weight in the directed connected graph, thereby generating a directed weighted connected graph; performing emotion perception identification on published contents to obtain an emotion value; and according to the directed weighted connected graph and the emotion value, obtaining relationship levels of the microblog users through a probability transfer model. The problem of low relationship level obtained in the prior art is solved, so that richer interpersonal relationships can be obtained.

The invention provides a vehicle-mounted interaction system based on emotion perception and voice interaction. The system is applied to a vehicle-mounted terminal and comprises an information acquisition module, an emotion analysis module and an execution module. The camera collects passenger information and transmits the passenger information to the analysis module, the emotional state of a passenger is obtained through a facial emotion analysis method, and the execution module starts matched preset operation. The preset operation comprises modes of intelligent voice interaction, automatic driving system starting, RGB atmosphere lamp adjustment, music adjustment, in-vehicle fragrance adjustment and the like. The emotion and intention of the passenger can be captured and analyzed in real time, interaction is carried out in a voice mode, the expectation of the user for intelligent voice and emotional perception is met, for the passenger, the change of the emotion can influence the change of the environment in the vehicle, the change is natural and comfortable, the experience of the passenger in the vehicle can be improved, and the user experience is improved. And natural interaction between people and vehicles is realized.

The invention discloses an information recommendation method and device based on emotion perception and a storage medium. The method comprises the steps of obtaining behavior data of a user; performing feature extraction on the behavior data to obtain an emotion feature vector; inputting the emotion feature vector into a trained information recommendation model to obtain recommendation information; wherein the information recommendation model is obtained by training based on the corresponding relationship between the emotion feature vector and the recommendation information. According to the method, the behavior data of the user is acquired, the equipment is controlled based on the behavior data, the content suitable for the emotion of the user is recommended to the user, the facial expression of the user does not need to be monitored, the conversation content of the user does not need to be analyzed, and the privacy safety of the user is protected.

The invention discloses a suicide emotion perception method based on multi-mode fusion of voice and micro-expression. The method comprises the following steps: collecting videos and audios by using Kinect with an infrared camera; analyzing image frames and audios in the video by using different methods and converting the image frames and audios into corresponding feature texts; fusing the featuretexts, namely performing dimension reduction processing to obtain fused features; and classifying the fusion features by using a SoftMax activation function, and judging whether the emotion belongs tosuicide emotions or not. According to the invention, the multi-modal data is aligned with the text layer. The text intermediate representation and the proposed fusion method form a frame fusing speech and facial expressions. According to the invention, the dimensionality of voice and facial expressions is reduced, and two pieces of information are unified into one component. The Kinect is used for data acquisition, and the invention has the advantages of being non-invasive, high in performance and convenient to operate.

The invention provides a music continuous emotion feature analysis evaluation method based on Gamma distribution analysis. The method includes the steps: firstly, building a Gamma distribution analysis evaluation method of music continuous emotion features, and finding emotion features most similar to emotion responses on a time sequence by the method; secondly, building an emotion feature analysis method based on an emotion perception matrix, evaluating the features from emotion perception ability by the method, and finding emotion features with best perception ability; finally, automaticallyanalyzing music emotions in real time based on Gamma distribution emotion analysis forecasting method. According to the method, the music emotions can be automatically analyzed, emotion tags are automatically forecasted, bases are provided for evaluation and selection of the music emotions, and the method has promoting effects on aspects such as artificial intelligence and emotion perception.

The invention provides a gait recognition and emotion perception method and system based on intelligent acoustic equipment. The method comprises the following steps of: taking a loudspeaker as a wavesource for sending signals, and utilizing a microphone to collect a signal reflected by a target pedestrian to obtain voice frequency data; processing the voice frequency data to obtain a corresponding energy atlas, and segmenting a signal containing a gait event; for the signal containing the gait event, independently extracting macroscopic gait characteristics, microcosmic gait characteristics and the embedding representation characteristics of various neural networks; and taking the blending of the characteristic vectors of the macroscopic gait characteristics, microcosmic gait characteristics and the embedding representation characteristics of various neural networks as input, and utilizing a trained classifier to obtain thee emotion classification result of the target pedestrian. Themethod can be applied to the intelligent home environment, and can carry out emotion recognition under a situation that individual privacy is not invaded and a user is not required to carry additionalequipment.

The invention provides a method for establishing an emotion perception model based on individual face analysis in a video, comprising the following steps: evaluating the emotional states of a plurality of tested individuals through a positive and negative emotion scale, and respectively obtaining positive emotion and negative emotion scores corresponding to the emotional states of the tested individuals; collecting tested face video data, wherein the tested face video data corresponds to the tested emotional state score; performing data denoising preprocessing on the face key points in the acquired video data in a two-dimensional space; selecting facial feature points representing emotion perception by calculating a difference measurement variance between adjacent frames; carrying out feature extraction, feature dimension reduction and feature selection on the facial feature points, and optimizing a feature set through a classifier by adopting a sequence backward selection (SBS) algorithm; and adopting a regression algorithm in machine learning and taking the obtained scores of the individual positive emotion and negative emotion as annotation data to perform model training and verification, thereby obtaining and storing a prediction model for the individual positive emotion and negative emotion. According to the method, the user does not need to report by himself/herself, thetimeliness is high, and the correlation coefficient of the model prediction score and the scale evaluation score can reach the medium-to-strong correlation level.

Disclosed is a hardware implementation method of an emotion perception function of an intelligent wearable device. The method includes the following steps: a PPG(photoplethysmography) sensor and an accelerometer transmit collected PPG signals and acceleration signals to a FPGA(Field-Programmable Gate Array) chip; the FPGA chip carries out feature extraction of the PPG signals and the accelerationsignals to obtain four physical quantities of heart rate, respiration rate, moving speed and moving frequency; and inputting the four physical quantities of heart rate, respiration rate, moving speedand moving frequency into an RHPNN(Robust Heteroscedastic Probabilistic Neural Network) based on probability to realize classification of emotion. The hardware implementation method and combination sense changes of emotion of a user in real time, remind the user to pay attention to metal health and adjust a working state in time, provide guarantee for mental health of the user, and can improve theworking efficiency of the user.

The invention discloses a speaker emotion perception method fusing multi-dimensional information, and relates to the technical field of deep learning and human emotion perception. The method includes: inputting a video of a speaker, and extracting an image and voice of the speaker from the video; inputting the image and the voice of the speaker into a multi-dimensional feature extraction network, extracting a language text and a language emotion in the voice, and extracting facial expression features of the speaker from image information; utilizing a multi-dimensional feature coding algorithm for coding various feature results of the multi-dimensional feature extraction network, and mapping the multi-dimensional information to a shared coding space; fusing the features in the coding space from low dimension to high dimension by using a multi-dimensional feature fusion algorithm, and obtaining feature vectors of multi-dimensional information highly related to the emotion of the speaker in a high-dimensional feature space; and inputting the fused multi-dimensional information into an emotion perception network for prediction, and outputting emotion perception distribution of the speaker. By means of the method, the ambiguity can be effectively eliminated according to the multi-dimensional information, and the emotion perception distribution of the speaker can be accurately predicted.

The invention provides a language emotion perception response method for a robot. The language emotion perception response method comprises the following steps that S1, a man-machine interaction module transmits collected information to a controller; S2, the controller transmits the collected information to an information analysis module; S3, the information acquired by the controller by the information analysis module is analyzed and disassembled, and disassembled data is transmitted to an emotion judgment module; S4, the emotion judgment module compares and judges the received information with the data of the storage module and then sends a judgment result to the controller; and S5, the controller controls the emotion expression module to express the corresponding emotion. According to the language emotion perception response method for the robot, facial expression and limb and sound expression can be carried out, so that emotion expression is more stereoscopic, emotion management ismore flexible, and people can relax in the chatting process.

The application relates to emotion adjustment method, device, computer equipment and a storage medium. The method includes obtaining the current emotional value and the expected emotional value of a user, wherein the current emotional value is obtained by calculating the physiological data of the user by using an emotional perception system. The music selection parameter is determined according tothe expected emotion value and the current emotion value, and the corresponding target music is selected from the preset music library according to the music selection parameter. Then the selected target music is played and the output emotion value of the user is calculated by the emotion perception system based on the played target music. When the difference between the output emotion value andthe expected emotion value is greater than the preset threshold, the output emotion value is used as the current emotion value for the next time. Then the method returns to step of obtaining the current emotional value and the expected emotional value of the user and repeats the step until the difference between the output emotion value and the expected emotion value is less than or equal to the preset threshold. The method can increase the emotion adjustment efficiency.

The invention relates to a virtual scene experience system with tactile simulation and emotion perception. The virtual scene experience system comprises a scene computer, and a display terminal and a touch simulation and feedback device which are respectively in communication connection with the scene computer; wherein the touch simulation and feedback device is worn on a limb of a user and comprises a microprocessor, and a second communication module, a positioning module, a skin tactile simulation module and a physiological signal module which are respectively connected with the microprocessor; the skin touch simulation module generates different pressures and shearing forces on the skin contacted with the skin touch simulation module according to the scene displayed by the display terminal and the limb position of the user, and simulates the touch of the user touching an object in the scene. The system provided by the invention provides a virtual scene for a user, accurately tracks the motion trail and posture of the hand of the user, simulates more vivid touch experience, can be widely applied to virtual simulation of various scenes, and greatly improves the user experience.