210 results about "Multiple classifier" patented technology

The voice print system of the present invention is a subword-based, text-dependent automatic speaker verification system that embodies the capability of user-selectable passwords with no constraints on the choice of vocabulary words or the language. Automatic blind speech segmentation allows speech to be segmented into subword units without any linguistic knowledge of the password. Subword modeling is performed using a multiple classifiers. The system also takes advantage of such concepts as multiple classifier fusion and data resampling to successfully boost the performance. Key word / key phrase spotting is used to optimally locate the password phrase. Numerous adaptation techniques increase the flexibility of the base system, and include: channel adaptation, fusion adaptation, model adaptation and threshold adaptation.

A method of identifying a malware file using multiple classifiers is disclosed. The method includes receiving a file at a client computer. The file includes static metadata. A set of metadata classifier weights are applied to the static metadata to generate a first classifier output. A dynamic classifier is initiated to evaluate the file and to generate a second classifier output. The method includes automatically identifying the file as potential malware based on at least the first classifier output and the second classifier output.

The invention discloses a system and a method for filtering sensitive webpage, which is based on multi-classifier fusion. The processing object is a webpage, and the processing result is whether the webpage contains sensitive content, which may be pornography, reaction, violence and other unhealthy Internet contents harmful to society. The system comprises a data stream obtaining and preprocessing unit, an image and text stream filtering unit and an information fusion unit of image filter and text filter, by the cooperation of multiple classifiers, the system acquires source code of a webpage by using the URL of the webpage, a text and an image are separated at preprocessing stage to obtain text information and effective image information; an input webpage is divided into three modes by decision tree algorithm; the webpage is recognized by using a consecutive text classifier, a discrete sensitive text classifier and an image classifier, the output result recognized by the classifiers is fused and calculated, then a judge factor is given, and the final result is returned to a browser.

The voice print system of the present invention is a subword-based, text-dependent automatic speaker verification system that embodies the capability of user-selectable passwords with no constraints on the choice of vocabulary words or the language. Automatic blind speech segmentation allows speech to be segmented into subword units without any linguistic knowledge of the password. Subword modeling is performed using a multiple classifiers. The system also takes advantage of such concepts as multiple classifier fusion and data resampling to successfully boost the performance. Key word / key phrase spotting is used to optimally locate the password phrase. Numerous adaptation techniques increase the flexibility of the base system, and include: channel adaptation, fusion adaptation, model adaptation and threshold adaptation.

The invention provides a universal circuit breaker mechanical fault diagnosis method based on feature fusion of vibration and sound signals. The method includes steps of 1, collecting machine vibration signals and machine sound signals during an engaging and disengaging process of a universal circuit breaker; 2, adopting an improved wavelet packet threshold value denoising algorithm for denoising; 3, adopting a complementary total average empirical mode decomposition algorithm for extracting a plurality of solid mode function components reflecting state information of engagement and disengagement actions of the circuit breaker from the denoising signals; 4, determining the number Z of the solid mode function components; 5, calculating the energy ratio, the sample ratio and the power spectrum entropy as three types of features; 6, adopting a combination core principal component analysis method for performing dimension reduction on a feature sample with unified three types of features of the vibration and the sound signals and obtaining M principle components; 7, establishing a related vector machine based sequence binary tree multiple classifier model.

Methods and apparatus for finding a match between a target bit pattern and multiple filter bit patterns. A filter array is created from the filter bit patterns and at least one intermediate array is generated from the filter array. Specific columns of the intermediate arrays are then extracted based on bit values of the target bit pattern. A row by row AND operation is performed on these columns to arrive at a match vector. the match vector identifies which of the filter bit patterns in the filter array match the target bit pattern. The method is implemented by using multiple classifier elements operating in parallel with each classifier element handling multiple filter bit patterns.

The invention relates to a multi-feature fusion phishing webpage detection method, which comprises two parts such as a training process and a detection process. The multi-feature fusion phishing webpage detection method integrates three views of phishing webpage characteristics by combining a semi-supervised learning tri-training method, and mainly solves a problem that the existing phishing webpage detection methods mostly need to perform classification model training by using supervised learning through a large amount of annotation data. The method provided by the invention mainly combines a coordinated training algorithm, starts from webpage URL characteristics, webpage information characteristics and webpage search information characteristics, applies the idea of multiple views and multiple classifiers to phishing webpage detection, and achieves the purposes of reducing the total numbers of manual annotation training samples and timely recognizing a phishing webpage through coordinated training and learning of different classifiers.

The invention discloses a deep neural network for the fine recognition of vehicle attributes and a training method thereof. The network comprises a depth residual network, a feature migration layer, aplurality of all-connection layers, a plurality of loss calculation units, and a plurality of parameter updating units. The depth residual network is used for carrying out feature extraction on an input image to obtain a feature image. The feature migration layer comprises a plurality of feature migration units and is used for enabling each of all feature migration units to be adapted to specifictasks according to the features shared by all attribute identifying tasks. The plurality of all-connection layers correspond to the branches of all attribute identifying tasks and are connected withthe feature migration layer so as to obtain feature vectors corresponding to all attribute identifying tasks. The plurality of loss calculation units correspond to the branches of all attribute identifying tasks and are respectively connected with the all-connection layers. The plurality of loss calculation units are used for calculating the loss of a loss function by adopting cross entropies as multiple classifiers. The plurality of parameter updating units correspond to the attribute identifying tasks and are connected with the loss calculation units. The parameter updating units are used for returning the loss based on the random gradient descent optimization algorithm, and updating parameters. According to the invention, various fine vehicle attributes can be identified at the same time by adopting only one neural network.

The disclosed multi-classifier combination face recognition method based on feature sorting comprises: extracting face area from initial image for pre-process and feature extraction; feature sorting to obtain different face feature groups; designing component classifier for every group to recognize face and combine results for optimal effect. This invention overcomes dimension disaster, reduces algorithm complexity, and improves recognition performance.

A body of text comprises a plurality of unknown attributes and a plurality of unknown values. A first classification sub-component labels a first portion of the plurality of unknown values as a first set of values, whereas a second classification sub-component labels a portion of the plurality of unknown attributes as a set of attributes and a second portion of the plurality of unknown values as a second set of values. Learning models implemented by the first and second classification subcomponents are updated based on the set of attributes and the first and second set of values. The first classification sub-component implements at least one supervised classification technique, whereas the second classification sub-component implements an unsupervised and / or semi-supervised classification technique. Active learning may be employed to provide at least one of a corrected attribute and / or corrected value that may be used to update the learning models.

The invention particularly relates to a sentiment dictionary learning based text sentiment analysis method and system. The method includes steps: subjecting initial text data to normalization processing to generate preprocessed text data, and clustering the preprocessed text data to a preset field; adopting a seed sentiment dictionary based sentiment word discovery method to form a special sentiment dictionary in the preset field; subjecting the preprocessed text data to retrieval according to the special sentiment dictionary to acquire target text data serving as initial training corpus in the corresponding preset field, and classifying inputted to-be-classified texts through formed multiple classifiers. Labor cost is reduced, the problem of overfitting caused by a single classifier is avoided, and accuracy in text sentiment analysis is improved by consideration of the text related fields.

The invention relates to an object detection method based on a cascade convolution neural network. The object detection method based on the cascade convolution neural network comprises steps of usinga convolution neural network to extract image characteristics, using a regional candidate network to generate a certain quantity of object candidate frames, using an optimized network to optimize a candidate framework, inputting the optimized object candidate frames into a detection network containing multiple classifiers to produce a preliminary detection result, using a binary classifier to perform re-detection on each kind of objects, and removing an error target to obtain a final accurate detection result. The object detection method based on the cascade convolution neural network uses anadvantage that the deep convolution network has a strong expression ability for a target, constructs a cascade convolution neural network for target detection, brings forward a new method of optimizing the target candidate frame and a strategy of eliminating an error detection sample, improves algorithm detection accuracy and can obtain a better target detection result.

The invention relates to a face identification method and a face identification system based on fusion of multiple classifiers. The method comprises following steps that a second classifer acquires a left-handed face image with a pose range of [-90, -15], a third classifier acquires a front face image with a pose range of [-15, +15], a fourth classifier acquires a right-handed face image with a pose range of [+15, +90], and comparing faces in the face images; fusing the face image with the comparison results of the same compared person in a pose module database to acquire an identification result, the identified relevant information of the compared person is displayed or the left-handed face image, the front face image and the right-handed face image are stored into the pose module database. Due to the adoption of the method and the system, a first comparison result, a second comparison result and a third comparison result can be effectively fused, the variation situation of the face poses in an application environment can be effectively processed in real time, and the accuracy and the robustness for identifying the multi-pose face can be improved.

The invention provides a rejection method for identifying handwritten character based on multiple classifiers, belonging to the field of handwritten character identification. The rejection method is led into a handwritten character identifying system, so as to effective improve the reliability of the identifying system and improve the identifying rate of the identifying system. The invention provides a rejection method based on a single classifier and the rejection method based on the multiple classifiers. Aiming at the limitation of a single characteristic rejection method, the idea of a multiple classifiers system is adopted to design the single classifier for the each characteristic, and integrate the rejection result of the each classifier, so as to play advantages of the each characteristic and improve the reliability of the rejection. Based on the rejection method of the single classifier, the invention provides the rejection method of the multiple classifiers. The method can better solve contradiction between the identifying rate and the rejecting rate of the handwritten character identifying system, and improve the reliability of the identifying system.

The invention discloses a fault detection method based on an analytic hierarchy process and weighted vote decision fusion. According to the method, firstly, multiple classifiers are selected as fusion sub-classifiers, multiple classifier models are established through a training data set, corresponding evaluation indexes are obtained according to the classification result of each classifier, and then multiple classifiers are scored and sequenced with the analytic hierarchy process, so that each classifier is endowed with the corresponding weight. Finally, multiple classifier decision results are integrated with a weighted vote fusion method, and the final fault detection result is obtained. Compared with other methods at present, the fault detection method has the advantages that the monitoring effect of the industrial process is improved, mastering and operation confidence of a process operator for the process are enhanced, the limitation of single fault detection methods is overcome to a great extent, and automatic implementation of the industrial process is better facilitated.

The present invention is method of recognizing and shielding the sex part in fetus image includes the following steps: 1. processing the original fetus image, finding out the edge contour of the sex part, and determining at least one initial locating point of suspected sex part; 2. recognizing the initial locating point by means of the fetus sex image database and recognizing system with multiple classifier fusion structure to judge the fetus sex part; and 3. shielding the fetus sex part with shielding image. The present invention makes it possible to avoid illegal fetus sex identification in ultrasonic apparatus effectively.

The invention discloses a multi-measurement time series similarity analysis method applicable to k-neighbor inquires of a time series. A multi-single-similarity-measurement method is chosen according to the analysis requirement, each single similarity measurement is used to analyze and inquire an m-neighbor sequence or subsequence of the sequence, pruning the m-neighbor sequence or subsequence under each similarity measurement to obtain a candidate similarity sequence or subsequence, and combining the candidate similarity sequence or subsequence by using a multiple-classifier combination method with advantage weight to obtain the k-neighbor sequence of the inquired sequence. Compared with the single similarity measurement, the similarity analysis of combined multiple measurements can obtain a more comprehensive analysis result. The multiple-classifier combination method with advantage weight regulates the ranking score according to the difference of the similarity distance between the adjacent candidate similarity sequence or subsequence and the inquired sequence while using a BORDA counting method for reference, so as to reflect the specific difference of similarity of the candidate similarity sequence or subsequence.

The invention belongs to the field of active driving, and particularly relates to a blind people detection and identification method and system based on combined characteristics and vehicle-mounted cameras. With the help of a vehicle-mounted sensor, on the basis of pedestrian detection within the forward visual range, detection is conducted on the detected pedestrians for the combined characteristics such as tactile sticks, guide dogs and glasses for blind people, the blind people are identified, and a driver is reminded to pay special attention to the blind people. Movement trends of the blind people and the guide dogs are judged according to attitude information of the blind people and the guide dogs, and the driver is reminded to better keep both the driver and the blind people safe. According to the blind people detection and identification method and system based on the combined characteristics and the vehicle-mounted cameras, by means of combination of multiple classifiers of offline trained pedestrians, the tactile sticks, the guide dogs and the glasses for the blind people, searching and detecting with a geometrical relationship are conducted on targets concurrently, and the efficiency and accuracy of detection are improved.

The invention provides a fuzzy integration multiple classifier integration-based uterine neck cell image identification method and device. The uterine neck cell image feature identification method comprises the following steps: in step S10, a single uterine neck cell image is subjected to preprocessing operation, edges of cytoplasm and cell nucleuses can be highlighted on the cell image, and a uterine neck cell image background can be removed; in step S20, the preprocessed uterine neck cell image is subjected to improved CV model segmenting operation; in step S30, the segmented uterine neck cell image is subjected to feature extraction operation and then subjected to dimension reduction operation via a genetic algorithm; in step S40, three single classifiers are integrated via fuzzy integration, and then the uterine neck cell image with reduced dimensions is subjected to identifying operation. The uterine neck cell image identification method provided in the invention can be used for precisely segmenting the cytoplasm and the cell nucleuses of uterine neck cells, segmenting speed can be increased, defects of low precision of one single classifier can be compensated by multiple classifier integration, and therefore uterine neck cell identification rate can be improved.

The invention discloses a high-resolution remote sensing image classifying method based on fusion of multiple classifiers. The high-resolution remote sensing image classifying method comprises the following steps that first, a training sample set is selected in an area of interest; second, the multiple classifiers are used for classifying remote sensing images; then, areas with the ground feature category classifying precision being lower than a threshold value a are classified again by using a voting method based on priori knowledge; at last, areas with the ground feature category classifying precision being lower than a threshold value b are classified by using a fuzzy decision template method, and finally the classified result of the target images is obtained. According to the high-resolution remote sensing image classifying method based on fusion of the multiple classifiers, the advantages of a single classifier are concentrated furthest, the disadvantages of the single classifier are restrained, the influences of 'same object with different spectrums ' and 'different objects with the same spectrum' on the classifying precision are lowered, and the precision of high-resolution remote sensing image classifying is improved.

The present invention provides a system and method for image steganalysis, relating to the fields of network transmission, information security, digital image processing, machine learning and the like. The method comprises a step of establishing an original image database, generating a steganographic image database for the image hidden information in the original image database, extracting the characteristics of the original image database and the images in steganographic images, and generating multiple sample spaces, a step of carrying out sample training on the multiple sample spaces, generating multiple classifiers, combining the multiple classifiers to generate a new classifier through a voting mode, and carrying out steganalysis on the image uploaded by the user through the new classifier. According to the system and the method, different steganographic methods can be selected to carry out steganography on the image, the characteristic of an image set is extracted and is converted into vector data, and different characteristic extracting methods can be employed at the same time.

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.

The invention relates to a multi-group image classification method based on two-dimensional empirical modal decomposition and wavelet denoising, belonging to the filed of image processing. The invention aims at solving the problems of insufficient utilization of image essential characteristics and low classification precision of the traditional classification method. The method comprises the following steps of: firstly, respectively carrying out two-dimensional empirical modal decomposition on each wave band in multi-group images to obtain the former K two-dimensional components and one residual error; secondly, summarizing the former K two-dimensional components as a characteristic value, and obtaining a denoised characteristic value after wavelet denoising; thirdly, randomly and proportionally selecting the denoised characteristic values of a plurality of multi-group images as training samples and test samples of a support vector machine, carrying out parameter training of the support vector machine on the training samples, and then carrying out attribution judgment to form a plurality of sub-classifiers of the support vector machine; and fourthly, constructing multiple classifiers based on a one-to-one strategy by utilizing the sub-classifiers of the support vector machine, and determining the attribution classes of the test samples according to a strategy function to complete the classification of the multi-group images.

A MMR system that uses multiple classifiers for predicting, monitoring, and adjusting index tables for image recognition comprises a plurality of mobile devices, a pre-processing server or MMR gateway, and an MMR matching unit, and may include an MMR publisher. The MMR matching unit includes a plurality of recognition unit and index table pairs corresponding to classifiers to be applied to received image queries, as well as an image registration unit for storing and monitoring performance data for the classifiers. The MMR matching unit receives the image query and identifies, using a classifier set, a result including a document, the page, and the location on the page corresponding to the image query. The present invention also includes methods for monitoring online performance of a multiple classifier image recognition system, for classifier selection and comparison, and for offline classifier prediction.

The invention provides a sign language recognition translation method based on the Davinci technology. The method comprises the following steps that an action video of a signer is collected; skin color blocks are extracted from a foreground image; the depth information of the extracted skin color blocks are calculated, and the face area is positioned according to the depth information of the skin color blocks; the hands are positioned according to the depth information of the skin color blocks, and key frames are extracted and the skin color blocks are repaired and filled; one or more hand shape features of the key frames are extracted; according to the extracted hand shape features, a single classifier is used for classification or multiple classifiers are used for voting, and a static gesture recognition result is obtained; the movement track of the gesture is recognized, and a track recognition result is obtained; the static gesture recognition result and the track recognition result are corrected and spliced, and a sign language recognition translation result is obtained. The sign language recognition translation method based on the Davinci technology has the advantages of being high in recognition rate and real-time performance.

The invention discloses an industrial process fault detection method based on multiple classifiers and the D-S evidence fusion. The industrial process fault detection method includes the steps of firstly, independently and repeatedly sampling normal data in the industrial process, then building corresponding classifier models for new training model data with multiple classifier methods, integrating and combining multiple classifier decisions through the D-S evidence theory, and obtaining final monitoring results. Compared with other methods at present, the industrial process fault detection method has the advantages that the monitoring effect in the industrial process can be greatly improved, the delay detection time can be shortened, the monitoring performance is improved to a great degree, the comprehensive ability and the operation confidence of a process operator to the process are improved, and the industrial process can be easily and automatically implemented.

The invention relates to an increment-based naive Bayesian network multiple classifier integration method, comprising the following steps that: a integration classifier and various key parameters are initialized; if no novel data exists, the process is ended; a category of a novel data item is forecasted by utilization of the prior integration classifier; parameter values of all the individual classifiers are dynamically updated; the weighing of all the individual classifiers is updated; if no error of the category forecast of novel data by the integration classifier is generated, all the individual classifiers in the integration classifier are trained by utilization of the novel data item; redundant individual classifiers are deleted according to the KL pruning strategy; a novel individual classifier is increased; all the individual classifiers are trained by utilization of the novel data item. The increment-based naive Bayesian network multiple classifier integration method can effectively improve the classification forecast result when concept shift is generated, and is particularly suitable for processing the concept shift problem.

The invention discloses an image meaning parsing method based on soft glance learning and aims to realize segmentation of images into a series of complete areas having individual meanings and meaning marking on each area on a basis of a lot of given user marking images. The image meaning parsing method comprises steps that: dual joint spectrum clustering and discrimination clustering methods are utilized to cluster image sub-areas acquired through an over-segmentation method, moreover, a corresponding constraint relation between image grade marking and image area grade marking is utilized to establish a soft glance learning model aiming to have a minimum error, and meaning labels are distributed to clustering sets in each image sub-area. Multiple classifiers acquired through discrimination clustering learning can realize meaning parsing for images without label information. The image meaning parsing method can not only add meaning labels to the images, but also the labels can be added to corresponding areas of the images to realize finer image meaning understanding.