1673results about How to "Reduce training time" patented technology

The invention discloses a fundus image retinal vessel segmentation method and a fundus image retinal vessel segmentation system based on deep learning. The fundus image retinal vessel segmentation method comprises the steps of performing data amplification on a training set, enhancing an image, training a convolutional neural network by using the training set, segmenting the image by using a convolutional neural network segmentation model to obtain a segmentation result, training a random forest classifier by using features of the convolutional neural network, extracting a last layer of convolutional layer output from the convolutional neural network, using the convolutional layer output as input of the random forest classifier for pixel classification to obtain another segmentation result, and fusing the two segmentation results to obtain a final segmentation image. Compared with the traditional vessel segmentation method, the fundus image retinal vessel segmentation method uses the deep convolutional neural network for feature extraction, the extracted features are more sufficient, and the segmentation precision and efficiency are higher.

A centralized dispute resolution system for use in commercial credit or debit card transactions. The inventive system utilizes a common data storage, interfaces, processes, procedures, rules, and other elements of a dispute resolution system that are made available to cardholders, merchants, card issuers, payment processors, and other parties that may be involved in a dispute or in a process intended to resolve a dispute. The system may be implemented using a client-server architecture with communication between clients and one or more server elements provided by a communications network.

The invention discloses a Chinese named entity recognition method based on BERT-BiGRU-CRF. The method comprises three stages: in the first stage, preprocessing mass text corpora, and pre-training a BERT language model; in the second stage, preprocessing the named entity recognition corpus, and encoding the named entity recognition corpus through the trained BERT language model; and at the third stage, inputting the encoded corpus into a BiGRU+CRF model for training, and performing named entity recognition on the to-be-recognized statement by using the trained model. Construction of the Chinesenamed entity recognition method based on BERT-BiGRU-CRF is carried out, semantic representation of characters is enhanced through a BERT pre-training language model, semantic vectors are dynamicallygenerated according to contexts of the characters, and the ambiguity of the characters is effectively represented. Compared with a method based on fine tuning of a language model, the method has the advantages that training parameters are reduced, and the training time is saved.

The invention discloses a voiceprint identification method. The voiceprint identification method comprises the following steps of: 1, preprocessing segmented speech data of each speaker in a training speech set to form a group of sample sets corresponding to each speaker; 2, extracting Mel-frequency cepstrum coefficients from each sample in all sample sets; 3, selecting a sample set one by one and randomly selecting the Mel-frequency cepstrum coefficients of part samples of the sample set, and training a Gaussian mixture model for the sample set; 4, performing incremental learning on the samples which are not selected and trained in the step 3 and the Gaussian mixture model of the sample set corresponding to the sample set one by one to obtain all optimized Gaussian mixture models, and optimizing a model library by utilizing all optimized Gaussian mixture models; and 5, inputting and identifying test voice data, identifying the Gaussian mixture model of the sample set corresponding to the test voice data by utilizing the optimized model library in the step 4, and adding the test voice data to the sample set corresponding to the speaker.

A body vibration apparatus includes an at least partially rigid platform, a first motor coupled to the platform such that movement of the first motor imparts force to the platform. The first motor has a first shaft that rotates a first eccentric weight in a first direction, phase and plane. A second motor is coupled to the platform such that movement of the second motor imparts force to the platform. The second motor has a second shaft parallel to the first shaft that rotates in a second direction, which, in one embodiment, is opposite the first direction. A second eccentric weight is coupled to the second shaft in the first plane. The second eccentric weight rotates with the second shaft at the first phase.

The invention discloses a method and a system for authenticating a shielded face, wherein the method comprises the following steps: S1) collecting a face video image; S2) preprocessing the collected face video image; S3) performing detection calculation on the shielded face, evaluating a position of a face image by utilizing a three-frame difference method according to motion information of a video sequence, and further confirming the position of the face according to an Adaboost algorithm; and S4) performing authenticating calculation on the shielded face, dividing a face sample into a plurality of sub-blocks, performing shielding distinguishment on the sub-blocks of the face by adopting a SVM(Support Vector Machine) binary algorithm combined with a supervising 1-NN k-Nearest neighbor method, if the sub-blocks are shielded, directly abandoning the sub-blocks, and if the sub-blocks are not shielded, extracting a corresponding LBP (Length Between Perpendiculars) textural feature vector for performing weighting identification, and then using a classifier based on a rectangular projection method to reduce feature matching times. According to the method for authenticating the shielded face, the detection rate and the detection speed for the local shielded face are effectively increased.

A noninvasive brain computer interface (BCI) system includes an electroencephalography (EEG) electrode array configured to acquire EEG signals generated by a subject. The subject observes movement of a stimulus. A computer is coupled to the EEG electrode array and configured to collected and process the acquired EEG signals. A decoding algorithm is used that analyzes low-frequency (delta band) brain waves in the time domain to continuously decode neural activity associated with the observed movement.

The invention discloses a traffic state prediction method for the urban road network based on key road sections, which is characterized by comprising the steps of first, carrying out data preprocessing; second, establishing a spatial weight matrix of the road network; third, establishing a time correlation matrix; fourth, recognizing key road sections by using a time-space correlation matrix; andfifth, establishing a deep convolution neural network, predicting the state of the road network in the future, and carrying out evaluation on a prediction model. The traffic state prediction method predicts the urban traffic flow state from a level of the wide-range road network, thereby being conducive to guiding the traffic flow from a macroscopic perspective, and fully exploring time-space correlation characteristics of the traffic flow. The key road sections in the road network are recognized, so that the training time of the model can be greatly reduced compared with a method of taking historical states of all road sections as input data, and the prediction efficiency is improved; and the convolution neural network is adopted to serve as the prediction model, and the prediction resultis also more accurate.

The invention discloses a diagnosis method for the fault position and the performance degradation degree of a rolling bearing, belonging to the technical field of fault diagnosis for bearings, and solving the problems of low accuracy of diagnosis for fault position and performance degradation degree, and high time consumption of training existing in an intelligent diagnosis method for a rolling bearing in the prior art. A white noise criterion is added in the disclosed integrated empirical mode decomposition method, so that artificial determination for decomposition parameters can be avoided, and the decomposition efficiency can be increased; and via the disclosed nuclear parameter optimization method based on a hypersphere centre distance, the small and effective search region of nuclear parameters in a multi-classification condition can be determined, so that training time is reduced, and the final state hypersphere model of a classifier is given. The intelligent diagnosis method based on parameter-optimized integrated empirical mode decomposition and singular value decomposition, and combined with a nuclear parameter-optimized hypersphere multi-class support vector machine based on the hypersphere centre distance is higher in identification rate compared with the existing diagnosis method. The diagnosis method disclosed by the invention is mainly applied to intelligent diagnosis on the fault position and the performance degradation degree of the rolling bearing.

The invention discloses an on-line sequential extreme learning machine-based incremental human behavior recognition method. According to the method, a human body can be captured by a video camera on the basis of an activity range of everyone. The method comprises the following steps of: (1) extracting a spatio-temporal interest point in a video by adopting a third-dimensional (3D) Harris corner point detector; (2) calculating a descriptor of the detected spatio-temporal interest point by utilizing a 3D SIFT descriptor; (3) generating a video dictionary by adopting a K-means clustering algorithm, and establishing a bag-of-words model of a video image; (4) training an on-line sequential extreme learning machine classifier by using the obtained bag-of-words model of the video image; and (5) performing human behavior recognition by utilizing the on-line sequential extreme learning machine classifier, and performing on-line learning. According to the method, an accurate human behavior recognition result can be obtained within a short training time under the condition of a few training samples, and the method is insensitive to environmental scenario changes, environmental lighting changes, detection object changes and human form changes to a certain extent.

A method and system for managing a media advertising enterprise including process and workflow capabilities for enterprise data matching. An EDM (Enterprise Data Management) module can be configured to include a set of rules at an enterprise level to manage disparate and disconnected records associated with an entity. A number of unmatched and enterprise entities that matches with respect to an active entity can be returned based on a fuzzy logic. A matching process can then be performed to accurately match the active entity and the unmatched entities with respect to a parent enterprise entity. The unmatched entity can be put on hold if additional information is required for performing a right match after assigning the parent enterprise entity. A note can also be added in order to place the unmatched entity on hold. Such an optimization mechanism can interactively manage and report records at the enterprise level in a simple and efficient manner.

The invention discloses a small data set face recognition algorithm based on machine vision, which comprises the following steps: 1) construction and pretreatment of the data set; 2) generating a virtual image for a category with insufficient images by using a generated antagonistic network; 3) carrying out data expansion on all kinds of images by using a data enhancement algorithm; 4) using convolution neural network to construct a model suitable for data; 5) training the design model; 6) using a camera to obtain a target image, importing the trained model, and carrying out face recognition.The algorithm of the invention can carry out face recognition with high accuracy for application scenes lacking face data.

The invention provides a sentiment classification method for generating a model deep-convinced-degree network on the basis of the probability of depth study. According to the technical scheme of the method, a plurality of Boltzmann machine layers are stacked, namely, output of this layer is used as input of the next layer. By the adoption of the mode, input information can be expressed in a grading mode, and abstraction can be conducted layer by layer. A multi-layer sensor containing a plurality of hidden layers is the basic study structure of the method. More abstract high layers are formed through combining the characteristics of lower layers and are used for expressing attribute categories or characteristics, so that the distribution type character presentation of data can be discovered. The method belongs to monitoring-free study, and a mainly-used model is the deep-convinced-degree network. The method enables a machine to conduct characteristic abstract better so as to improve the accuracy of sentiment classifications.

The invention discloses a gesture recognition method based on deep learning. The gesture recognition method comprises the following steps: carrying out noise reduction processing to a collected gesture image, and eliminating a color offset phenomenon in the gesture image; locking an area, where a gesture is positioned, in the image by adopting a frame difference method and a color characteristic detection method, and tracking a gesture by adopting a CamShift algorithm to obtain a gesture target; carrying out the deep learning on a gesture target image; and inputting the obtained gesture image to be recognized into a trained deep belief network model to finish the recognition and the classification of the gesture.

The invention discloses a Chinese domain term recognition method based on mutual information and a conditional random field model. The Chinese domain term recognition method includes the following steps: (1) gathering domain text corpus and marking all the punctuations, spaces, numbers, ASSCII (American Standard Code for Information Interchange) characters and characters except Chinese characters in the corpus; (2) setting character strings and computing the mutual information values of the character strings, (3) computing the left comentropy and the right comentropy of every character string, (4) defining character string evaluation function, setting evaluation function threshold, computing the evaluation function values of every character string, determining that every character string is a word, comparing in sequence the evaluation function value of the former character with the evaluation function value of the latter character in the character string and segmenting character meaning character strings one by one, (5) utilizing conditional random fields to train a conditional random field model and recognizing domain terms with the conditional random field model. When the Chinese domain term recognition method is used to recognize terms, the data sparsity of legitimate terms is overcome, the amount of calculation of conditional random fields is reduced, and the accuracy of the Chinese domain term recognition is improved.

The invention relates to a robust ear detection method which improves an AdaBoost arithmetic and belongs to the technical field of image mode identifying. The invention is characterized by proposing an ear detection method with excellent performances under a complex underground. The invention proposes four anisomerous Haar-like corner characteristics which are used for describing the grayscale changes on the partial areas of the ears; a policy of subsection selection is adopted for selecting the best sorting threshold of the Haar-like characteristics, thus reducing the sample training time; the weight of a weak sorter is modified for reducing the mistaken detection rate of the sorter; the threshold HW is set and eliminated according to the distribution change of the sample weight in the training, thereby preventing an over-studying phenomenon from being generated and leading the miss-detection rate and the mistaken direction rate of the ear detection to be reduced; besides, the invention also provides a single-ear detection policy for leading both the defection efficiency and the detection effect to be improved. The excellent performances of the robust ear detection method are shown on a PC machine and a DSP.

The invention discloses a speaker recognition system and a speaker recognition method. The speaker recognition system comprises a characteristic extraction unit, a background model generation unit, a registered speaker model generation unit, a metric value calculation unit and a recognition unit, wherein the characteristic extraction unit is configured to extract a characteristic vector of speech data of a speaker; the background model generation unit is configured to perform internal clustering on the characteristic vector of the speech data of a background speaker and generate a universal background model aiming at a normal speaker according to the result of the internal clustering; the registered speaker model generation unit is configured to adapt to the universal background model by using the characteristic vector of the speech data of each registered speaker so as to generate a registered speaker model of each registered speaker; the metric value calculation unit is configured to calculate metric values of the characteristic vector of a tested speaker on the universal background model generated by the background model generation unit and on the registered speaker model of each registered speaker, which is generated by the registered speaker model generation model; and the recognition unit is configured to recognize the tested speaker according to the metric values calculated by the metric value calculation unit.

The embodiment of the invention provides an image foreground and background segmentation network model training method, an image foreground and background segmentation method, an image processing method and device, and a terminal device. The image foreground and background segmentation network model training method comprises the following steps: obtaining the eigenvectors of a sample image to be trained; performing convolution on the eigenvectors to obtain the convolution results of the eigenvectors; magnifying the convolution results of the eigenvectors; determining whether the magnified convolution results of the eigenvectors satisfy the convergence conditions or not; if so, completing the training of the convolutional neural network model used for segmenting the foreground and background of the image; and if not, adjusting the parameters of the convolutional neural network model according to the convolution results of the amplified eigenvectors and performing iteration training on the convolutional neural network model according to adjusted parameters of the convolutional neural network model until the convolution results satisfy the convergence conditions. By means of the image foreground and background segmentation network model training method, the training efficiency of the convolutional neural network model is improved, and the training time is shortened.

The invention provides an image identification method based on depth learning and transfer learning. The image identification method based on depth learning and transfer learning includes the following steps: 1) a preparation step: reading a pre-training model, reading an image directory, and dividing a training set, a verification set and a test set; 2) a training step: constructing a full connection neural network classifier, taking the image set as input of the pre-training model, and using the output of the pre-training model to update the full connection neural network classifier; and 3) a storage step: storing the model result. The image identification method based on depth learning and transfer learning has the advantages of integrating with the application of depth learning and transfer learning so as to provide a relatively accurate bladder cancer diagnosis result for a user, on the basis of the extremely limited training time and training sample quantity.

The invention proposes a crack identifying method based on deep learning convolutional neural network, and the method comprises the following specific steps: A) collecting an image building, training and verifying database; B) establishing a deep learning convolution neural network; and C) using the image database in A to train the completed deep learning neural network in B.

The invention belongs to the technical field of image processing and mode recognition and particularly relates to a license plate recognition method based on a deep convolutional neural network. The method includes: performing license plate detection on vehicle images, performing image segmentation on detected license plates to obtain license plate characters, using the license plate characters as training samples to obtain a training sample block set, inputting the training sample block set into a deep auto-encoder to train the deep auto-encoder, using the trained deep auto-encoder as the convolution kernel of the convolutional neural network, extracting the convolution features of the training sample block set, performing pooling operation on the convolution features of the training sample block set to obtain feature vectors, performing normalization processing on the feature vectors, loading the feature vectors after the normalization processing into an SVM classifier to train the SVM classifier, and testing to-be-recognized vehicles. By the method, license plate recognition accuracy can be increased, and license plate character recognition rate and robustness can be increased when the license plate characters are located in severe environments.

The invention discloses a neural network model training method, device and system and belongs to the image processing field. The method comprises the following steps that: medical samples sent by a plurality of clients are received, wherein the medical sample sent by the first client comprises a plurality of CT images and first label images corresponding to each CT image, wherein the first label images are used for identifying specified organs contained in the CT images, the first client adopts a local neural network model to segment the plurality of CT images so as to form the first label images, and the first client is any one client in the plurality of clients; and a first neural network model in a server is trained according to the medical samples sent by the plurality of clients, wherein the first neural network model is a neural network model of the latest version in the server. With the neural network model training method, device and system of the invention adopted, the training time of the neural network model is shortened, and the accuracy of the training of the neural network model is effectively improved. The neural network model training method, device and system of the present invention are used for training neural network models.

The invention discloses a residual error learning-based CT sparse reconstruction artifact correction method and system. The method comprises the steps of firstly, reconstructing sparse projection datagenerated by CT through an FBP (Filtered Back-Projection) method, wherein a image after reconstruction has a serious artifact; secondly, learning features of the artifact by establishing a residual error neural network structure to obtain an artifact image; and finally, recovering a clear CT image by performing residual error operation on the sparse reconstruction image and the artifact image. Based on sparse reconstruction, a residual error learning-based convolutional neural network structure framework is introduced; by adopting GPU acceleration, the training time is shortened, the trainingprocess of an experiment is accelerated, and very high CT reconstruction image quality can be realized at a very low CT projection angle; and therefore, the unnecessary injury of radiation to a humanbody is effectively reduced.

The invention belongs to technical fields characterized by protocols and discloses an attack occurrence confidence-based network security situation assessment method and system. According to the attack occurrence confidence-based network security situation assessment method and system, a machine learning technology is adopted to analyze network stream data and calculate a probability that networkstreams belong to attack streams; a D-S evidence theory is used to fuse the information of multi-step attacks to obtain the confidence of attack occurrence; and a network security situation is calculated by means of situational factor integration on the basis of security vulnerability information, network service information and host protection strategies; and therefore, the accuracy of assessmentis effectively improved. Since the confidence information of detection equipment is added to the assessment system, the influence of false negatives and false positives can be effectively reduced. Anensemble learning method is adopted, so that the accuracy of confidence calculation can be improved. A network attack is regarded as a dynamic process, and merging processing is performed on the information of the multi-step attacks. Information fusion technology is adopted, so that network environment characteristics such as vulnerabilities, service information and protection strategies are comprehensively considered.

The invention discloses an online deep learning SLAM based image cloud computing method. The image cloud computing method comprises the following steps: acquiring image data and storing the image data; extracting a key frame and uploading the key frame; using the image data to construct a data set and training the data set to obtain optimal convolutional neural network parameters; extracting real-time image feature points and performing recognition, and performing feature point matching on adjacent frame images; iterating the image feature points to obtain the best matching transformation matrix, performing correction by using position and pose information, and obtaining the camera pose transformation; obtaining the optimal pose estimation through registration of point cloud data and the position and pose information; transforming the pose information into a coordinate system through matrix transformation, and obtaining map information; repeating the previous steps in regions with insufficient precision; and allowing a client to display the result and performing online adjustment at the same time. The invention parallelizes image processing, deep learning training and SLAM by usingthe cloud computing technology to improve the efficiency and accuracy of image processing, positioning and mapping.

The invention discloses a remote sensing image rapid target detection method based on a deep Hash auxiliary network, and solves the problem that an existing target detection framework cannot carry outefficient detection on a background area which does not have a target and occupies a large area in a wide remote sensing image. The method comprises the following implementation steps: generating anexperimental training image block set and a test image block set; constructing a remote sensing image rapid target detection model based on a deep Hash auxiliary network; training a remote sensing image rapid target detection model based on a deep Hash auxiliary network; and using the trained deep Hash auxiliary network to obtain a test result of the test image block set, and counting the runningtime. According to the method, the Hash auxiliary branch network module is introduced, redundant calculation in target detection is greatly reduced by quickly filtering most non-target background areas in the wide remote sensing image, the detection time of the non-target background areas is greatly shortened, and the detection efficiency is greatly improved. The method is used for target detection of wide remote sensing images.

A feature automation process defines step-by-step instructions for involving automation engineers, defining, implementing and reviewing software test automation during the development of a feature or product. This process seamlessly integrates the roles of automation engineers and other resources into the software development life cycle (SDLC). An enterprise first creates a dedicated automation team. The feature automation team preferably works with a product / feature team to enable the latter team to better understand the roles of the automation engineers and to further facilitate transparency into the product / feature requirements, design and implementation activities. The feature automation process enables a quality assurance (QA) team to offload (to the feature automation team) the responsibility of writing test scripts, and for creating an automation framework, test designs, and for implementing and maintaining test code. The process ensures that all stakeholders are involved in the reviewing the automation framework and test design prior to test implementation to enhance the reusability of the framework and the stability of the test runs.

The invention relates to a license plate character recognition method based on real-time vehicle tracking and binary index classification, which comprises the following steps of: dynamically and continuously performing multi-point tracking on a vehicle in real time, and recognizing the license plate character based on a least squares support vector machine (LS-SVM) and the binary index classification. A novel license plate-extracting scheme which comprehensively adopts gray gradient, shape and posture, vision models and the like is provided in the stage of segmenting and extracting the license plate character by using the space distribution information of the license plate character. The robustness and accuracy of extracting the license plate are improved, and the real-time property is guaranteed.

The invention discloses a short-term power load prediction method based on a GRU neural network and transfer learning, and belongs to the field of power load prediction. A short-term power load prediction problem is solved. The invention provides a short-term power load prediction method based on a GRU neural network and transfer learning. A GRU-based recurrent neural network is utilized to solvethe problem that a traditional neural network cannot extract time characteristics, auxiliary information input influencing load change is expanded, including influence factors such as date, temperature and weather, a Dropout layer and a normalization layer are added to avoid the overfitting problem, and the accuracy of load prediction is improved; Historical knowledge is migrated through migrationlearning, a network which is normally put into use is adjusted, retraining and fine tuning are carried out through target prediction data, the value of the historical data is brought into play, and the precision and efficiency of load prediction are further improved.

The invention discloses a furniture layout method based on a part-by-part reinforcement learning technology. The furniture layout method comprises a furniture layout environment building step, a furniture layout reinforcement learning training step and a furniture layout generation step by using reinforcement learning. Firstly, evaluation and scoring are conducted on a specific furniture layout scheme through the manual technology, and feature extraction processing is conducted on the data; secondly, regression learning is conducted through a neural network algorithm, and the neural network obtained through training is used for simulating designers to score; secondly, learning specific state space and behavior action space by using a reinforcement learning technology according to environment feedback as guidance; and finally, in actual use, the trained reinforcement learning model is used for carrying out layout on specific furniture. The method is high in applicability in the implementation process, automation of furniture layout is achieved, the design cost is reduced, and the design efficiency is extremely high.