264 results about "Person recognition" patented technology

The invention discloses a conference recording method based on the Internet of Things, and relates to the technical field of conference recording. A microphone array and a direction sensor are adopted to position a spokesman, a wearable microphone is combined to obtain individual voice features, voice is matched with the identity of the spokesman, and in the spokesman recognition process, the voice is recognized through a voice recognition module; the speech streams of different spokesmen are distinguished in combination with a speech round detection technology, a speech time axis is generated, for speech transcription, background noise is removed through a noise reduction technology, meanwhile, semantic compensation is introduced, and texts are corrected in combination with projection and whiteboard writing data collected by IoT equipment, so that the recognition accuracy is improved; the method comprises the following steps: additionally carrying out topic segmentation, converting a text into a high-dimensional vector by adopting a semantic vectorization method, determining a topic boundary by utilizing similarity calculation and a clustering algorithm, carrying out dynamic adjustment in combination with IoT position data and individual voice features, accurately classifying the same discontinuous topic discussion, and preventing cross speaking of different topics from being mistakenly classified into a whole.

The invention relates to the technical field of computer vision and cross-visual-angle pedestrian recognition, in particular to a cross-visual-angle pedestrian re-recognition method based on visual language prompt learning, and the method comprises the steps: obtaining a target image; the target image is input into a preset pedestrian recognition model, a pedestrian re-recognition result is output, the pedestrian recognition model is obtained by training a visual language pre-training model CLIP through a prompt learning mechanism and a two-stage training strategy, the prompt learning mechanism is used for modeling visual angle deviation, and the two-stage training strategy is used for training visual angle deviation. The double-stage training strategy is used for realizing cross-modal semantic alignment. According to the invention, the accuracy and robustness of cross-view identification can be significantly improved.

The embodiment disclose a wearable augmented reality eyewear system including a frame supporting left and right transparent lenses. A plurality of outward-facing cameras mounted on the frame have overlapping fields of view for capturing real-time images of the environment. A plurality of inward-facing cameras integrated in the frame or lenses detect pupil size, pupil dilation, and gaze direction using infrared imaging. A digital processor analyzes gaze direction and fixation based on virtual grid coordinates, initiates an object or person recognition routine using images from the outward-facing cameras, and compares detected features with records in a database. At least one lens includes an AR projection system embedded therein and configured to superimpose data about the recognized object or person in the user's field of view. An audio speaker positioned near the user's ear outputs related audio. A wireless communication module enables real-time data exchange with a cloud-based platform.

A method includes obtaining image frames capturing one or more people in at least one scene and identifying features of the image frames. The method also includes providing the identified features to a trained spatiotemporal transformer machine learning model configured to generate a set of features for each of the one or more people. The set of features for each person includes facial features of the person and pose features of the person over time. The method further includes performing face identification using the facial features to generate one or more first embeddings representing at least one face of at least one person and performing gait identification using the pose features to generate one or more second embeddings representing at least one gait of at least one person. In addition, the method includes identifying at least one of the one or more people based on the first and second embeddings.

The invention discloses a zero-configuration adaptive speaker recognition method and system, and relates to the technical field of voice signal processing. The method comprises the following steps: receiving an audio stream, carrying out voice activity detection, obtaining a single-person voice segment, extracting a voiceprint embedding vector of the single-person voice segment, and carrying out online clustering to generate a speaker identity pool; and the identity pool is updated by calculating the multi-dimensional fusion similarity between the remaining single-person voice segments and the voice model, and temporary identity tags corresponding to the single-person voice segments are output. According to the speaker recognition method provided by the invention, a voiceprint template does not need to be registered in advance, and the flexibility and real-time performance of the system are improved.

A system includes a set of crawlers that find and retrieve documents from an information network, an information extraction system, a knowledge graph storing nodes and edges that connect them, wherein each node represents a respective entity of a corresponding entity type of a plurality of entity types, and wherein the knowledge graph further stores event data relating to events detected by the information extraction system, a machine learning system that trains models that are used in connection with at least one of entity extraction, event extraction, recipient identification, and content generation, a lead scoring system that scores the relevance of information to an individual and references information in the knowledge graph, and a content generation system that generates content of a personalized message to a recipient who is an individual for which the lead scoring system has determined a threshold level of relevance.

The invention relates to a workpiece identity control method for an iron tower component steel plate tailor-welding robot, and belongs to the technical field of welding automation and intelligent manufacturing. S2, workpiece clamping and positioning control; s3, performing three-dimensional identification and identity confirmation; s4, the workpiece state and the program are subjected to self-adaption; s5, dynamically correcting the welding path; s6, process real-time monitoring and alarming are carried out; and S7, process tracing and database archiving are conducted, accurate recognition and identity confirmation of workpieces are achieved by constructing a steel plate component three-dimensional model database and combining image recognition and 3D scanning, a welding program and a path template are automatically matched based on a recognition result, weld joint deviation is dynamically corrected, and the welding quality is improved. According to the method, the recognition accuracy, the path tracking precision and the process stability of the welding robot can be remarkably improved, manual intervention is reduced, and the automatic high-precision welding production requirements of multi-model and multi-batch iron tower components are met.

The invention discloses a conference summary generation method based on voice recognition, relates to the technical field of conference summary generation, and solves the technical problem that the recognition precision of a voice recognition technology is reduced due to the influence of noise in the prior art. The method comprises the following steps: acquiring initial data of enterprise employees and a plurality of types of noise data; respectively training based on the initial data and the noise data to obtain a voice transcription model and a character recognition model; voice data in the real-time conference are transferred based on the voice transfer model; training based on the historical conference content text to obtain a correction model; correcting the real-time conference content text based on a correction model to obtain a target conference content text; recognizing a spokesman in the voice data based on a character recognition model to obtain identity information of the spokesman; and matching the target conference content text with the identity information of the spokesman to obtain a target text, and inputting the target text into the long text cognition large model to obtain a conference summary, so that the technical problem is solved.

The invention discloses a speech enhancement training method and device based on speaker perception, equipment and a medium, and relates to the technical field of artificial intelligence, and the method comprises the steps: inputting a first noisy speech sample into an initial speech enhancement joint training system, so as to enable a pre-training speech enhancement module to carry out the denoising of the first noisy speech sample, outputting a first enhanced voice sample, and determining voice enhancement loss; extracting a first speaker sensitive feature of the first enhanced voice sample through a pre-training sharing encoder, carrying out speaker ID classification on the first speaker sensitive feature through a pre-training speaker identification module to obtain a speaker ID classification prediction result, and calculating a speaker identification loss; and judging whether the sensitive features of the first speaker have noise or not through a confrontation training discriminator, and adjusting confrontation loss according to a noise judgment result, so as to update system parameters based on each loss, and obtain a trained target speech enhancement combined training system. And the conflict between the two tasks of speech enhancement and speaker recognition is coordinated.

The invention provides a speaker recognition method and system, a terminal and a medium, and the method comprises the steps: constructing and training a voice feature extraction depth model through employing a first time-frequency fusion residual network, and constructing a voice feature extraction lightweight model through employing a second time-frequency fusion residual network; and taking the voice feature extraction depth model as teacher network training to obtain a convergent voice feature extraction lightweight model, so as to further perform feature extraction on the to-be-recognized voice data subjected to logarithm Mel spectrum feature extraction to generate a corresponding to-be-recognized voice feature vector, comparing the target speech feature vector with each template speech feature vector in a speech library, and determining a target speech feature vector and a target speaker; according to the speech feature extraction lightweight model, the calculation cost of the model and the occupied equipment memory are greatly reduced, the frequency domain information and the time domain information in the speech data to be recognized can be comprehensively and accurately captured, and the speech feature extraction lightweight model has high recognition precision and strong anti-noise capability.

The invention discloses a group key person identification method, electronic equipment and a readable storage medium, and belongs to the technical field of mobile communication. According to the method, the target sub-graph is sent to the terminal equipment corresponding to the target entity, so that the terminal equipment carries out fusion based on the received target sub-graph, and a global heterogeneous network graph corresponding to the target entity after multiple knowledge graphs are linked is obtained; training a local business model according to the global heterogeneous network atlas to obtain local training gradient data, and sending the local training gradient data to a service party; the service party fuses the local training gradient data provided by each service party to obtain global gradient data corresponding to the graph federal learning model, and feeds back the global gradient data to each service party; and respectively updating model parameters of a local service model by each service party according to the fused global gradient data. According to the invention, the group key person identification precision and efficiency can be improved.

The invention relates to a limited space operation safety monitoring method based on multi-sign parameter intelligent judgment, which comprises the following steps of: training a single-class support vector machine model by utilizing historical data, and establishing a normal physiological state multi-dimensional reference model under each operation content; the method comprises the following steps: acquiring current operation content in a limited space, determining a vital sign parameter set needing to be monitored, and collecting real-time vital sign data of an operator through a user terminal; calculating a comprehensive abnormal score AS of the real-time vital sign data, and when the comprehensive abnormal score AS exceeds a dynamic threshold value delta, marking the data as comprehensive abnormal data; marking the vital sign parameter with the maximum contribution degree absolute value in the abnormal parameters as a key vital sign; and sending alarm information to a monitoring terminal, wherein the alarm information comprises the identifier of the key vital sign. By comprehensively analyzing the real-time vital signs and comparing the real-time vital signs with the dynamic threshold value, the accuracy of abnormal person recognition is effectively improved.

A vehicle includes multiple storage zones. A set of imaging sensors defines fields of view and each storage zone is included in at least one defined field of view. A controller is in communication with the set of imaging sensors and includes a set of processing modules configured to cooperatively implement a smart storage tracking system including a person recognition processing module and an object recognition processing module. The person recognition processing module includes a process for identifying unique persons using image analysis and the object recognition module includes a process for identifying unique objects using image analysis. The smart storage tracking system defines an object to person and location data entry in a database using the person recognition processing module and the object recognition processing module. The database is in communication with the controller and storing each defined object to person and location data entry.

Methods and systems for a borrower identification and prediction of credit risk associated with the identified borrower is provided. The system includes a server that retrieves an electronic message that includes details of transactions conducted by a first entity over a time. The electronic message is parsed based on a predefined parameter, to identify a second entity, where the first entity has conducted a business transaction with the second entity. Further, a first set of details associated with the second entity is obtained from one or more resources associated with the server. Furthermore, a knowledge graph is generated based on the business transaction and the first set of details. The knowledge graph indicates an association between the first entity and the second entity. Subsequently, a credit risk score associated with lending money to the second entity is predicted based on the knowledge graph.

In accordance with an aspect of the present invention, a crime prevention, monitoring and alerting device is provided. The crime prevention, monitoring and alerting device includes a personal protection device having a refillable ampule unit with a button, a plurality of sensors, an audio-visual capturing module, a storage device a sound unit and a microcontroller. The crime prevention, monitoring and alerting device then includes a central server with a machine learning unit and an artificial intelligence unit. The crime prevention, monitoring and alerting device further includes a communication unit. In an embodiment, the communication unit to communicate data to security group and emergency contacts, identify presence of anyone at the geo location and notify about the crime, receive request to identify safest route and provide safest route based on time, location, crime, presence of security group to the user.

The present invention relates to a training method, device, equipment and medium for a pedestrian recognition model, belonging to the field of artificial intelligence technology, and the method includes: constructing a semantic feature encoder integrated with an attention mechanism; extracting semantic features of a sample data set through the semantic feature encoder; parsing the sample data set to obtain body shape features; obtaining KL divergence loss and feature difference loss based on the semantic features and the body shape features; performing weighted calculation based on at least KL divergence loss and feature difference loss to obtain model loss parameters; training a pedestrian recognition model to be trained according to the model loss parameters until the training converges to obtain a trained pedestrian recognition model. The scheme of the present invention extracts the semantic features of the original pedestrian more fully by integrating a pixel-by-pixel attention mechanism into the backbone network, and suppresses the interference of clothing-related information on the model by learning the identity information of the pedestrian through the body shape branch, so that the model pays attention to deep-level clothing-irrelevant feature information.

The embodiment of the invention belongs to the field of multi-camera systems, and relates to a multi-target area dynamic vehicle and pedestrian recognition method, which comprises the following steps: acquiring image data of a target area acquired by a plurality of cameras; preprocessing the image data, and extracting features of the target area; target detection is carried out on the image data, vehicle and pedestrian targets in the image data are identified, and the positions and features of the targets are determined; performing preliminary positioning on each target, and predicting the motion trail of the target through a Kalman filtering algorithm or a particle filtering algorithm; and carrying out space-time association on the targets identified by the plurality of cameras based on a joint data association method of the positions, the features and the motion trails of the targets. The invention also provides a multi-target area dynamic vehicle and pedestrian identification system. The coverage rate and accuracy of target identification are improved.

According to one embodiment, an information processing device for a gesture-based ordering taking system or the like includes a processor. The processor is configured to detect a person making a particular gesture in image data from a camera, identify each person making the particular gesture in the image data as a first type person, then identify each first type person meeting a candidate condition in the image data as a candidate. The processor then designates a candidate as a subject according to a subject selection condition, and then detects gestured-based input from the subject in additional image data from the camera after the subject has been designated.

The invention discloses an anti-lost supervision method and system for elderly people, and relates to the technical field of intelligent pension, and the system comprises a positioning module and a far-end background. Wherein the positioning module comprises a packaging shell, a controller, a positioning chip, a communication module and a pressure switch, the controller, the positioning chip, the communication module and the pressure switch are integrated in the packaging shell, the positioning module is used for being installed in a shoe sole, and the sensing side face of the pressure switch faces the foot sole of a user; the far-end background is configured to receive positioning data; when a position query request initiated by the personal terminal is received, outputting matched positioning data based on the electronic map and the requester identification information; judging whether the user is in a return state or not based on the matched positioning data; if the current positioning data is in the return state and the user staying time exceeds the preset time, calling the current positioning data to carry out difference comparison analysis with the preset home address positioning data, and outputting a comparison result; and outputting a preset matching prompt based on the comparison result. The method has the effect of improving the user experience.

The application discloses a low-light target detection method based on a double-layer optimization theory, which significantly improves the enhancement quality and detection accuracy. The method first adopts a double-layer learning architecture to bridge the enhancement and detection tasks, and organically combines the low-light image enhancement task and the target detection task. Secondly, the enhancement network parameters are taken as the main optimization target, and the detection network parameters are taken as the constraint condition. The enhancement network generates images that are conducive to machine perception. Finally, in order to improve the training efficiency, a method based on the implicit function theory is proposed to approximately solve the corresponding gradient. The training cost is effectively reduced, and the training speed is improved. The application significantly improves the target detection accuracy in weak light scenes, and can be applied to various target detection tasks in weak light scenes, such as night pedestrian recognition, automatic driving, security monitoring and the like.

The present application relates to the technical field of speech signal processing, and discloses a small sample speaker recognition method and system based on self-supervised speech representation, which comprises: multi-layer self-supervised speech representation extraction and inter-layer adaptive weighted fusion, channel-aware dimension alignment projection, time-channel cross-attention feature learning of a double-flow Transformer encoder, confidence-aware dynamic prototype aggregation and speaker classification, and classification confidence-driven representation refinement feedback; the present application fully utilizes the hierarchical information of the self-supervised model through multi-layer representation fusion, excavates discriminative features from the time and channel dimensions through the double-flow cross-attention mechanism, and introduces a confidence-driven closed-loop feedback mechanism to improve the recognition accuracy of low-confidence samples.

The invention relates to an actuation system (2) for an adjustable element (2) of a motor vehicle, comprising: • a person identification device (8) which is designed to authenticate and identify an operator (4); • a gesture recognition device (9) which is designed to recognize an operating gesture of the operator (4), • a control device (10) which is designed to generate a control command as a result of a signal from the gesture recognition device (9); • an automatic mechanism (11) which is designed to adjust the element (2) depending on the control command of the control device (10), wherein an ultrahaptic device (8) which is designed to generate a virtual ultrasonic wave element (2) which can be haptically perceived by the operator (4) and on which the actuation gesture can be performed by the operator (4). Furthermore, the invention relates to a method for actuating an adjustable element (2) of a motor vehicle by means of such an actuating system (2).

A method and system for use of obfuscation coding. An example method includes a computing system receiving an image that depicts a human face. The method then includes the computing system generating a modified image based on the received image, with the generating of the modified image involving inserting into the received image an obfuscation marker that (i) obfuscates the human face in a manner that prevents a person observing the modified image from identifying the human face and (ii) includes predefined symbology that is interpretable by a machine to identify the obfuscated human face. Further, the method includes using the predefined symbology in the modified image as a basis to take action based on an identity of the human face.

Systems and methods are directed to redacting and reinstating personal identifying information (PII) from text data. A PII management system accesses an input text and identifies, using one or more redaction components, PII mentions in the input text to be redacted. A placeholder manager of the PII management system replaces each unique PII string of a final set of PII mentions with a non-PII string to generate redacted text, whereby the non-PII string is generated by the placeholder manager. The placeholder manager also generates a mapping dictionary that maps each unique PII string to the non-PII string that replaces it. The mapping dictionary is used to reinsert one or more unique PII strings after processing of the redacted text. The redacted text is then transmitted to a downstream component for the processing.

The invention discloses a face recognition control method for an elevator. The face recognition control method comprises the steps that images and depth information of passengers in an elevator entrance area are collected through an image collecting module; the image processing module processes the image, extracts face features of each passenger and compares the face features with information in a database; the behavior recognition module judges whether each passenger walks towards the elevator and enters the elevator or not according to the depth image and the posture estimation; the system establishes an association binding relationship between the face recognition and the entering behavior; an elevator call and a target floor selection instruction are sent to the passengers which are successfully recognized and have the authority; and the identification and behavior records are stored in a database for background audit analysis. The invention belongs to the technical field of artificial intelligence buildings, and particularly provides an artificial intelligence building system which is used for solving the problems that visitors are inconvenient to pass, recognition of multiple persons is confused, and follow-up potential safety hazards exist in the prior art.

A method and system for use of obfuscation coding. An example method includes a computing system receiving an image that depicts a human face. The method then includes the computing system generating a modified image based on the received image, with the generating of the modified image involving inserting into the received image an obfuscation marker that (i) obfuscates the human face in a manner that prevents a person observing the modified image from identifying the human face and (ii) includes predefined symbology that is interpretable by a machine to identify the obfuscated human face. Further, the method includes using the predefined symbology in the modified image as a basis to take action based on an identity of the human face.

The present invention aims to provide a program that allows each user to easily identify candidates who meet their desired referral criteria (candidate criteria). [Solution] The program of the present invention causes a referral system comprising at least a memory unit and a processing unit to execute a personnel memory function that stores user information, candidate information, and referrer information in the memory unit; a condition memory function that stores candidate conditions desired by one user, which are included in the user information of each user, in the memory unit; and a specification function that, if a candidate who meets the candidate conditions exists and a referrer who can introduce that candidate has a connection with the one user, identifies one or more such candidates as one or more suitable candidates to introduce to the one user.

A portable electronic device to be used as a keyless device of a vehicle includes a controlling circuit which includes a physiological feature detection module and a processing circuit. The physiological feature detection module is used for detecting a physiological feature of a person. The processing circuit is coupled to the physiological feature detection module and used for determining whether the person is an authorized user of the vehicle according to the detected physiological feature of the person, and for opening a car door of the vehicle or starting the vehicle when the person is identified by the processing circuit as the authorized user.

A method and system are provided for training a robot to recognize objects in the workspace of the robot. Objects in the workspace are identified by the user, and the robot determines candidate objects. Feedback may be used in order for the user to confirm whether the candidate object determined by the robot system matches the object intended by the user. Gripping information for the object may also be identified by the user to train the robot how to grip the object.