Patents
Literature
Hiro is an intelligent assistant for R&D personnel, combined with Patent DNA, to facilitate innovative research.
Hiro

618 results about "Data Annotation" patented technology

Data annotation is the task of labelling any type of data : images, audio, text, video, …. Generally, it is done by selecting a “zone” of the data, and adding a label to this specific zone.

Systems and methods for data annotation, recordation, and communication

InactiveUS20080186255A1Enhance tacticalEnhance strategic situation awarenessCathode-ray tube indicatorsInput/output processes for data processingSensory FeedbacksGraphics
Systems, devices, and methods to provide tools enhance the tactical or strategic situation awareness of on-scene and remotely located personnel involved with the surveillance of a region-of-interest using field-of-view sensory augmentation tools. The sensory augmentation tools provide updated, visual, text, audio, and graphic information associated with the region-of-interest adjusted for the positional frame of reference of the on-scene or remote personnel viewing the region-of-interest, map, document or other surface. Annotations and augmented reality graphics projected onto and positionally registered with objects or regions-of-interest visible within the field of view of a user looking through a see through monitor may select the projected graphics for editing and manipulation by sensory feedback from the viewer.
Owner:ADAPX INC

Method and apparatus for omni-directional image and 3-dimensional data acquisition with data annotation and dynamic range extension method

A method and apparatus for omni-directional image and 3-dimensional data acquisition with data annotation and dynamic rage extension method is capable of omni-directionally photographing, acquiring 3-dimensional images photographed by cameras having each different exposure amount in connection with the direction of height of an object, extending dynamic range, and generating an geographical information by entering an annotation such as photographing location and time into the photographed images, which can be connected with other geographical information system database. The apparatus includes one or more multi camera module(s) which are stacked and formed multi layers in the direction of height for acquiring 3-dimensional images and extending dynamic range of the 3-dimensional images, wherein each multi camera module includes a plurality of cameras symmetrically arranged with a specific point in a plane.
Owner:KUJIN LEE

Process for Automatic Data Annotation, Selection, and Utilization.

Systems and methods for the automatic annotation of data are disclosed, particularly a process and system for enabling users to generate automatic annotations, to select one or more of those annotations, and to utilize the selected annotations and their various relationships to the annotated data. Also disclosed are ways to utilize selected online annotations in the form of promotional material, so that the user can manage the online promotional material he / she selects, and obtain various advantages therefrom.
Owner:GLOBALINGUIST

Exstensibility application programming interface and framework for meta-model objects

The present invention relates to a system and methodology providing an Application Programming Interface (API) and framework that supports a meta-object model for application design and operating system interactions. The API includes an input component that receives data and / or instructions. The input component processes a meta-data model that interrelates data via a class hierarchy, wherein the class hierarchy includes relationship descriptions between class objects, attributes, rules, and / or behavioral descriptions. Another aspect includes a data management system that includes a component to receive an item having meta-data annotations. An analysis component determines at runtime a structure for the item via deployment of the meta-data annotations. The system can also include a framework component that defines meta-data class derivations, meta-data classes, meta-data integrity rules, and / or meta-data class behavior.
Owner:MICROSOFT TECH LICENSING LLC

Process for automatic data annotation, selection, and utilization

Systems and methods for the automatic annotation of data are disclosed, particularly a process and system for enabling users to generate automatic annotations, to select one or more of those annotations, and to utilize the selected annotations and their various relationships to the annotated data.
Owner:GLOBALINGUIST

Point cloud data annotation method, segmentation model determination method, target detection method and related equipment

The invention relates to a point cloud data annotation method, a segmentation model determination method, a target detection method and related equipment. The annotation method comprises the steps of obtaining original point cloud data collected by a laser radar; projecting the original point cloud data to a two-dimensional graph, and obtaining each piece of target annotation information based on the two-dimensional graph; detecting the original point cloud data through a trained target detection model, and determining detected target detection information; and on the basis of the corresponding target category confidence, adopting target detection information to correct the target annotation information, obtaining a target detection data set of the original point cloud data, wherein the target detection data set comprises the position information and the target category of each target. By adopting the method, marking can be rapidly and accurately carried out.
Owner:CHANGSHA INTELLIGENT DRIVING INST CORP LTD

Method and apparatus for managing multimedia files

A system for locating, segmenting, annotating and retrieving multimedia files, provides a database of metadata (68) relating to multimedia files (70), a database manager (64), and a database client (66) for accessing the data contained within the database. The database client (66), together with the metadata database (68) and database manager (64) provide a variety of different functionalities, namely a deep linking functionality, a segmentation functionality, a metadata annotation functionality, a retrieval functionality, and an access functionality. The user, through database client (68) annotates the multimedia file or segment of multimedia file with metadata which is saved in the database (68). When the user desires to locate a multimedia file, the metadata is searched or browsed to locate the database entry associated with the multimedia file in question. The system further provides the ability for the user to load, open and start the playing of a multimedia file from a point intermediate to its start and end point, and without the user having to experience the portion between the start point and the intermediate point, break down larger files into shorter segments or individual frames (or still images) or regions within frames (or still images) without creating additional copies of the selected segments of the multimedia files of interest, thereby avoiding copyright issues from arising.
Owner:TAGMOTION

A method for constructing emotion recognition model of Chinese social text based on deep fusion neural network

The invention discloses a method for constructing a Chinese social text emotion recognition model based on a deep fusion neural network, which comprises the following steps: data collection, using a Python Scrapy framework to construct a social text network crawler, and collecting picture and text data; Data preprocessing, preprocessing the Chinese text collected by the data acquisition module; Data annotation for emotionally annotating the processed text; Text vectorization, using Word2Vec tools to train the word vector; Model building, design integration BILSTM-CNN network model; Model training, will be marked after the text through BILSTM-CNN fusion neural network model for training. The invention constructs a deep fusion emotion analysis model, aiming at making full use of the featureextraction ability of the depth neural network model to carry out feature expression on Chinese emotion text, thereby constructing a multi-classification model of emotion, and improving the accuracy rate of automatic emotion multi-classification.
Owner:SOUTH CHINA UNIV OF TECH

System and method for annotating video with geospatially referenced data

A system that annotates video with geospatially referenced data on a display is provided. A video client is connected to a source of video data, the video client being capable of overlaying a video feed from the source of video data with graphics or text data not inherent in the video feed. An annotation provider receives a request for annotation from the video client and returns geographically relevant annotations for display to the video client. A coordinate transformation service device determines a geographic area in view in the video feed and provides accurate position information for security data contained in the video feed into display coordinates. The system may apply geo-spatial information to create a map of a structure in the video feed and to display annotations of locations in the area of field of view that are obstructed by the structure.
Owner:SIEMENS IND INC

Method, device and system for automatic recommendation of network content

ActiveCN103399891AEliminate manual data labelingSpecial data processing applicationsIncremental learningData Annotation
The invention discloses a method, a device and a system for automatic recommendation of network content. The method for automatic recommendation of network content includes receiving network text data and extracting keywords from the network text data, training a keyword classifier by an incremental learning process according to the keywords, receiving the network text data of specific users, classifying the specific users according to classification results of the keywords extracted from the network text data of the specific users by the aid of the keyword classifier, and recommending network content, interested by the specific users, to the specific users according to the classification of the network content predefined on the basis of the classification of the specific users. The method, the device and the system for automatic recommendation of the network content have the advantages that recommended site categories can be changed according to changes of personal interests of the users, and a great quantity of manual data annotation during training of the classifier is omitted.
Owner:BAIDU ONLINE NETWORK TECH (BEIJIBG) CO LTD

A data annotation crowdsourcing platform system based on an intelligent contract and a crowdsourcing data annotation method

The invention relates to the technical field of block chains, intelligent contracts, data crowdsourcing, machine learning and deep learning, in particular to a data annotation crowdsourcing platform system based on an intelligent contract and a crowdsourcing data annotation method. The crowdsourcing data annotation method comprises the steps of separating the data crowdsourcing annotation task from the annotation result quality evaluation task, evaluating annotation data through the trained model, randomly selecting a plurality of annotation data quality inspectors for crowdsourcing evaluation, and determining the data annotation quality by integrating a plurality of evaluation modes. The whole process of the crowdsourcing task is written into the block chain through the block chain infrastructure, the mutual trust of all parties participating in crowdsourcing is achieved, the integrity and the non-tamper performance of behaviors are guaranteed, the data labeling quality is improved toa certain degree, and then the model training efficiency and the model prediction accuracy are improved.
Owner:JINAN INSPUR HIGH TECH TECH DEV CO LTD

A voice data annotation method and apparatus

The invention brings forward a voice data annotation method and apparatus. The voice data annotation method includes the steps of receiving voice data, performing voice recognition of the voice data to obtain a recognized text; obtaining a user-confirmed text after the user confirms the recognized text; and performing analyzing of the recognized text and the user confirmation text to extract automatic annotation features; and performing annotation of the voice data according to the automatic annotation features and a pre-constructed automatic annotation model. The method can solve problems existing in manual annotation, raises the efficiency of voice data annotation and reduces the cost.
Owner:IFLYTEK CO LTD

Training method, matting method, device, medium and terminal for image processing model

The embodiment of the application discloses a training method, a matting method, a device, a medium and a terminal of an image processing model. The training method of the image processing model includes obtaining a trichotomy of the original image; Generating a training sample set according to the original image and the trichotomy; The preset depth learning network is trained based on the training sample set to obtain an image processing model, wherein, the image processing model is used for labeling the original image to obtain a trichotomy. By adopting the technical proposal, the input original image can be automatically annotated to obtain a trichotomy, and a large amount of hairline level data annotation can be carried out without manual annotation, the annotation workload can be reduced, and the image annotation efficiency can be improved. In addition, the matting effect can be optimized by using image processing model to label the original image, avoiding the error that may be introduced by manual labeling.
Owner:GUANGDONG OPPO MOBILE TELECOMM CORP LTD

Relation extraction and knowledge graph construction method based on deep learning model

The invention discloses a relation extraction and knowledge graph construction method based on a deep learning model. The relation extraction and knowledge graph construction method specifically comprises the following steps of processing the corpus data into the annotation data by utilizing a data annotation platform; inputting the annotation data into the deep learning model for training and testing, and generating a relationship extraction model; inputting the to-be-extracted text data into the relationship extraction model, and extracting the entity relationship pairs of the to-be-extracted text data; and constructing a knowledge graph of the to-be-extracted text by utilizing a graph database. According to the method, the relationship extraction model is constructed based on a BERT model, a bidirectional long-short-term memory network model and a conditional random field algorithm, the entity relationship pairs can be accurately extracted, and the correctness of the entity relationship pairs entering the graph database is ensured by manually checking the extracted entity relationship pairs.
Owner:DATAGRAND TECH INC

Modeling and controlling method for synchronizing voice and mouth shape of virtual character

ActiveCN108447474AEfficient natural lip-sync controlEfficient natural synchronization controlSpeech recognitionSpeech synthesisAttitude controlSynchronous control
The invention belongs to the virtual character attitude control in the field of speech synthesis, and particularly relates to a modeling and controlling method for synchronizing the voice and the mouth shape of a virtual character. The object of the invention is to reduce the mouth shape animation data annotation amount and to achieve accurate and naturally smooth mouth motion synchronized with the voice. The method comprises: generating a phoneme sequence corresponding to the to-be-synchronized voice; converting the phoneme sequence into a phoneme category sequence; converting the phoneme category sequence into a static mouth shape configuration sequence; and converting the static mouth shape configuration sequence distributed on a time axis into dynamically changing mouth shape configuration by a dynamic model; rendering the dynamically changing mouth shape configuration into an attitude image of the head and neck of the virtual character, and displaying the attitude image in synchronization with a voice signal. The method can realize efficient and natural virtual character mouth shape synchronous control without mouth shape animation data and with a phonetic prior knowledge anddynamic model.
Owner:北京灵伴未来科技有限公司

Synchronous video and data annotations

A video surveillance system captures video of a protected area, stores the video in a computer readable memory, stores data annotations in the computer readable memory where the data annotations are searchable using a search criteria, and stores links that link the stored data annotations to corresponding video segments of the stored video so that the data annotations can be used to search for a video segment of interest. The data annotations are searched using the search criteria, and the video segment of interest linked to the data annotation found as a result of the search is displayed.
Owner:COLUMBIA PEAK VENTURES LLC

Automatic video advertisement detection method

The invention belongs to the technical field of computer video processing and particularly relates to an automatic video advertisement detection method. The method comprises the following steps: a digital television signal or a local video format broadcast by a television station is input; then, shot segmentation, key frame extraction, FMPI frame classification, copy detection and suspected advertisement detection are carried out on a video by comprehensively adopting color, texture and edge features; and finally an advertisement component in the video is analyzed out. The automatic video advertisement detection method can be used for filtering advertisements in digital television rebroadcast and solving the problem of automatic video data annotation in projects. The method adopts the concept of FMPI key frame, greatly improves the advertisement copy detection accuracy and algorithm efficiency and provides a very reliable basis for making a classifier based on the characteristics of an FMPI key frame for suspected advertisement detection.
Owner:FUDAN UNIV

Semi-supervised learning-based method for filtering junk users in social network

ActiveCN106294590ASatisfy the requirement of conditional independenceReduce time complexityData processing applicationsWeb data indexingLearning basedData Annotation
The invention discloses a semi-supervised learning-based method for filtering junk users in a social network. A cooperative training algorithm is applied to the detection of the junk users in the social network. Massive information in the social network is classified mainly by utilizing a supervised learning algorithm at present; and the algorithm is based on a classification model built based on annotated data, but the social information scale is huge, the labor cost required for data annotation is high, and few methods for solving the problem of user data annotation of the social network exist. A method is proposed; and by referring to the cooperative training algorithm, multiple views and multiple classifiers are applied to a large amount of non-annotated social network data or a small amount of annotated social network data, so that the classifiers on different views learn mutually and the purpose of data annotation is achieved.
Owner:CHONGQING UNIV OF POSTS & TELECOMM

Data annotation method and terminal

The invention relates to a data annotation method a terminal. The data annotation method comprises the steps of obtaining an annotated data set Dall, wherein the data set Dall includes N pieces of data and annotation labels respectively corresponding to the N pieces of data, and N is a positive integer; dividing the N pieces of data into K parts so as to generate K first training samples, whereineach first training sample includes (K-1) parts of data, and the data in any two first training samples is not identical; training the data in each first training sample, and generating K first classification models; performing predicted annotation on data J by using the first classification model Mi, and determining a predicted label of the data J; and performing repeated annotation on the data Jif the predicted label of the data J is determined to be inconsistent with the annotation label. Therefore, the cost of performing repeated annotation on the annotated data is reduced, the time consumed by repeated annotation is reduced, the quality of data annotation is improved, and an excellent promotion role is played on improving the model effect.
Owner:TENCENT TECH (SHENZHEN) CO LTD

Method and device for management of data annotation

The invention provides a method and device for management of data annotation. The method comprises the steps that a dataset corresponding to a data annotation task is acquired, and annotation rules corresponding to various types of data in the dataset are acquired; the dataset is divided into a data subset; according to the acquired annotation rules corresponding to each type of data, data annotation sub-task description information of the data subset can be generated and released; the data subset is sent to a sender of a first claiming request aiming at the data subset; and annotated data from the sender of the first claiming request is received. The management method also comprises the steps that calling instructions for annotation tools corresponding to the each type of data in the data subset are sent; and / or the released data annotation sub-task description information of the data subset contains data formats of the various types of data in the data subset after the target annotation. The method and the device provided by the invention avoid data format conversion which is conducted to the annotated data.
Owner:ALIBABA GRP HLDG LTD

Data annotation, model training and image processing method and device and storage medium

The invention provides a lung focus detection algorithm based on weak annotation data. The algorithm comprises the steps of data labeling, model training and image processing, wherein a partially labeled weak data labeling method is adopted for labeling, a labeling result is perfected by referring to the thought of difficult case mining in the data labeling process, and it is guaranteed that a high-quality labeled data set is provided for subsequent model training under the condition that only part of focuses are labeled. According to the method, model training is combined with a 'weak data 'marking mode, positive and negative sample anchor points for network training are acquired from positive and negative sample sets respectively, and the focal loss hyper-parameters are adjusted according to the source characteristics of the positive and negative sample anchor points, so that the detection model obtains a better detection effect.
Owner:北京医准智能科技有限公司

Method of annotating portions of a transactional legal document related to a merger or acquisition of a business entity with graphical display data related to current metrics in merger or acquisition transactions

A method is provided for annotating portions of a transactional legal document related to a merger or acquisition of a business entity. An electronic data source is maintained that stores data related to a plurality of different metrics in merger or acquisition transactions. At least some of the data is updated as new merger or acquisition transactions occur. Portions of a legal document are electronically annotated with one or more annotations that graphically display data related to different metrics in merger or acquisition transactions. The annotations are created using the data stored in the electronic data source. The annotations reflect the most current data stored in the electronic data source. At least some of the annotations change over time as new merger or acquisition transactions occur.
Owner:SHAREHOLDER REPRESENTATIVE SERVICES

Method to display transaction performance data in real-time handling context

A system and method for providing a performance-based choreographer-architecture style visualization of an application program, in which an original electronic workflow representation of the application is modifying to incorporate certain run-time performance indicators as collected from a performance monitoring system. This provides a diagram, typically only available during the design phase, annotated with performance data typically only available in diagrams with other perspectives than the design perspective. The invention is especially useful for annotation workflow diagrams produced by choreographer architecture design systems.
Owner:IBM CORP

Visual compiler-based person and pose detection method

InactiveCN106897697ASolve the problem of vanishing gradientCharacter and pattern recognitionNeural architecturesPattern recognitionData Annotation
The invention provides a visual compiler-based person and pose detection method. The method mainly comprises the steps of performing data synthesis of scene description; learning a net through synthesized data; defining the net by adopting basic blocks; and performing joint locating by using a pose net. According to the process, the scene description is used as an input of a visual compiler, a pedestrian detection system and a pose estimation system are trained by calibrated real data annotations, and the net is learnt through the synthesized data; then the net is defined by using two basic units including a residual module and a spatial belief module; and finally pedestrians are located by using the pose net. According to the method, the detected annotations, body part positions and segmented mask codes can be automatically obtained, the pedestrians are located by using a video camera, and the pedestrians are subjected to pose estimation and activity analysis; and the influence of illumination, shielding and the like on the detection is reduced, so that the identification efficiency is effectively improved.
Owner:SHENZHEN WEITESHI TECH

Data annotation platform for expanding merchandise varieties and data annotation method

The invention discloses a data annotation platform for expanding merchandise varieties and a data annotation method. The platform comprises a task management module, a data annotation module, a data checking module and a model training module, wherein the task management module imports data to be annotated and an annotation mode; the data annotation module carries out data annotation on the data to be annotated to obtain annotated data; the data checking module checks the annotated data; and the model training module carries out model training on the checked annotated data to output a training model. When the data annotation platform is used, the problems that long time is consumed, and the accuracy rate cannot be easily ensured in the conventional data annotation mode can be solved.
Owner:BEIJING JINGDONG SHANGKE INFORMATION TECH CO LTD +1

Data annotation detection method and device

The invention discloses a data annotation detection method and device. The method comprises the steps: obtaining a target image and a manual annotation result obtained after the manual annotation of the target image, and the manual annotation result comprises the types and coordinates of a plurality of first objects; identifying the target image according to a predetermined target recognition model to obtain a target annotation result, the target annotation result comprising categories and coordinates of a plurality of second objects, and the target recognition model being obtained by trainingthe categories and coordinates of a plurality of sample objects obtained by annotating the sample image on the basis of a target detection network; According to the manual annotation result and the target annotation result, determining whether a wrong annotation result exists in the manual annotation result or not. Thus, since the accuracy of the recognition result of the target recognition modelis high, the target annotation result is used as a quality inspection standard, quality inspection personnel can be effectively helped to comprehensively detect wrong manual annotation results, and the accuracy of data annotation is improved.
Owner:北京云聚智慧科技有限公司

Large-scale farm sign abnormal bird detection system and detection method thereof

The invention relates to a large-scale farm sign abnormal bird detection method, which comprises the following steps: pre-processing the collected thermographic images for data annotation, then adopting an instance-based segmentation general framework (Mask RCNN) target detection and instance separation method for feature extraction, pixel alignment, target positioning, classification and mask separation for thermal imaging images, by controlling the target detection number of a single image, improving the detection accuracy rate in each image. The method is an attempt in the field of target detection of current thermal imaging images, breaks the limitations brought by the traditional methods, and uses the superior performance of deep learning in image feature processing to improve the robustness and accuracy of a model. The feasibility of the deep learning method in the field of thermal imaging image detection is verified. At the same time, the method can also be extended to the livestock and other fields, thus improving the level of intelligence in the breeding industry in China.
Owner:TIANJIN UNIV

Data annotation method, device, electronic device and storage medium

A data annotation method includes generating a data annotation model corresponding to each of a plurality of first data sets according to sample data in N first data sets and manual annotation of sample data in each of the first data sets; Determining an annotation of sample data in the second data set based on predictive annotation of sample data in the second data set predicted by N data annotation models and manual annotation of sample data in the second data set with the same proportion of predictive annotation less than or equal to a preset value; training N data annotation models by using sample data and corresponding annotations of the first data set and the second data set to obtain N trained data annotation models for data annotation. Embodiments of the present application can save labor costs by annotating data through data annotation models. Embodiments of the present application also provide a data annotation device, an electronic device, and a storage medium.
Owner:北京九狐时代智能科技有限公司
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Eureka Blog
Learn More
PatSnap group products