184results about How to "Improve selection accuracy" patented technology

Integrated intelligent system adapted for any operating system and / or multi-OS computing environment seamlessly having sensory input / data acquisition cum recording server group and / or analytics server group enabling fail-safe integration and / or optimized utilization of various sensory inputs for various utility applications. Also disclosed as added advancements include intelligent method / system for cost-effective and efficient band adaptive transferring / recording sensory data from single or multiple data sources to network accessible storage devices, fail safe and self sufficient server group based method for sensory input recording and live streaming in a multi-server environment, intelligent and unified method of colour coherent object analysis, face detection in video images and the like, resource allocation for analytical processing involving multi channel environment, multi channel join-split mechanism adapted for low and / or variable bandwidth network link, enhanced multi-colour and / or mono-colour object tracking and also an intelligent automated traffic enforcement system.

Genetically elite ungulate mammals are identified on the basis of gene expression profiles from biological samples such as liver and blood. Methods and compositions are presented to select genetically elite animals with a desired phenotype such as high milk production for breeding to improve production levels. A method to select an animal with a specific phenotype, e.g. milk production and health traits, includes creating a Gene Expression Index for a specific phenotype and using the index to identify candidate animals for breeding by comparing the index to gene expression profiles of the animals.

A control method of searching for a neighboring cell of a mobile station communicating with a base station is provided in a direct sequence CDMA mobile communication system which transmits information by carrying out double modulation using a first spreading code group and a second spreading code. The first spreading code group includes spreading codes that have a same repetition period as an information symbol period and are used in common by the base stations, and the second spreading code has a repetition period longer than the information symbol period. The base stations are assigned different second spreading codes. The control method stores at least one second spreading code and its phase into a first table, which second spreading code corresponds to a perch channel whose second spreading code and phase are known; stores a second spreading code used by a neighboring base station into a second table; searches for a perch channel whose second spreading code and phase are unknown; and searches for a perch channel whose second spreading code and phase are known. The neighboring cell search method can save the power consumption and time required for the mobile station to carry out the cell search with preventing an increase in the total cost of the system.

The invention relates to a method for presenting a number of two- or three-dimensional images from different modalities registered with each other, with points of interest being able to be assigned to an individual image of the modalities and in all images selectable graphics primitives being overlaid on the assigned points of interest so that a visual assignment of points of interest or area of interest between simultaneously displayed two- or three-dimensional images occurs, as well as an imaging system of a workstation for executing the method.

A laser module has an optical amplifier, a first etalon and a wavelength selectable mirror. The first etalon has wavelength peaks at a given wavelength interval in transmission characteristics and transmits a light from the optical amplifier. The wavelength peaks are tunable. The wavelength selectable mirror acts as an external mirror of an external cavity laser including the optical amplifier, and has a relatively high reflection intensity at a part of wavelength range in an effective gain range of the optical amplifier. The effective gain range of the optical amplifier includes more than one wavelength peak of the first etalon. A reflection bandwidth where the wavelength selectable mirror has a relatively high reflection intensity is less than twice of the wavelength interval of the wavelength peaks of the first etalon.

To reduce power consumption for scanning and maintain communication quality by reducing the time for scanning or connection when a wireless communication terminal switches between networks or base stations. A wireless communication terminal 1 enabling connection with a plurality of base stations disposed in different positions which includes a base station scanning unit 100 detecting and scanning a base station capable of communication at the current position of the wireless communication terminal, a simultaneous detection base station recording unit 101 recording simultaneously detected base stations as simultaneous detection information in a storage unit, a connection candidate extraction unit 102 and a first communication unit 103 performing communication by connection with the extracted connection candidate base station. Based on the simultaneous detection information, the connection candidate extraction unit 102 extracts a base station having the possibility of simultaneous detection with the currently connected base station as a connection candidate base station.

The present invention relates to a base station selection method based on deep reinforcement learning in an LTE-V. The method comprises the following steps of: 1) constructing a Q function according to LTE-V (Long Term Evolution-Vehicle) network communication features and base station selection performance indexes; 2) obtaining state information of a vehicle in a network through a mobile management unit, constructing a state matrix, and storing the state matrix into a playback pool; 3) taking an experience playback pool as a sample, and based on the constructed Q function, employing a dueling-double training mode to obtain a main DQN (Dueling-Double Deep Q Network) used to select an optimal access base station; and 4) performing processing of the input information by employing the main DQNobtained through training, and outputting a selected access base station. Compared to the prior art, the base station selection method based on deep reinforcement learning can consider the delay performance and the load balance performance of communication at the same time to allow the vehicle to timely and reliably perform communication, and is high in base station selection efficiency and highin accuracy.

Disclosed herein are genetic markers for animal growth, fatness, meat quality, and feed efficiency, methods for identifying such markers, and methods of screening animals to determine those more likely to produce desired growth, fatness, meat quality, and feed efficiency and preferably selecting those animals for future breeding purposes. The markers are based upon the presence or absence of certain polymorphisms in an HMGA nucleotide sequence.

A method of selecting a toric lens by taking into consideration the magnitude and orientation of the posterior cornea and / or the location of the incision axis is described. The magnitude and orientation of the posterior cornea can be calculated as a function of the measured pre-operative orientation of the steep meridian of the anterior cornea.

A scanning transmission electron microscope which enhances correction accuracy of a de-scanning coil for canceling a transmitted-electron-beam position change on an electron detector. Here, this transmitted-electron-beam position change appears in accompaniment with a primary-electron-beam position change on a specimen caused by a scanning coil. First, control over the scanning coil is digitized. Moreover, while being synchronized with a digital control signal resulting from this digitization, values in a de-scanning table registered in a FM (2) are outputted to the de-scanning coil. Here, the de-scanning table is created as follows: Diffraction images before and after activating the scanning coil and the de-scanning coil are photographed using a camera. Then, based on a result acquired by analyzing a resultant displacement quantity of the diffraction images by the image processing, the de-scanning table is created.

The invention discloses a cultivating method of a laying hen with low egg yolk cholesterol content. The cultivating method of the laying hen with the low egg yolk cholesterol content comprises the following steps: 1, using deoxyribonucleic acid (DNA) sequence containing very low density lipoprotein receptor (VLDLR) gene of the laying hen as a template, designing a primer, and using polymerase chain reaction(PCR) to restriction fragment length polymorphism (RELP) method to conduct sample hen individual genotyping; 2, using a gene effect analysis model to determine select and remain genotype; 3, establishing a family of cock and hen individuals conforming to the select and remain genotype; 4, extracting DNA of offspring of hens of a select and remain genotype family, using the primer in the step 2 to conduct PCR-RFLP analysis, and selecting and remaining hens of the select and remain genotype in the step 2; 5, repeating the step 3 and the step 4 subculture and seed selection at least three generations, and obtaining laying hens with the low egg yolk cholesterol content. The cultivating method of the laying hen with the low egg yolk cholesterol content adopts a VLDLR gene molecular marker assisted select breeding method, egg yolk cholesterol contents of the cultivated laying hens can decrease more than 20 percentages compared with the original level.

The invention discloses a surface-mounted type permanent magnet synchronous motor rotor initial position estimation method. Firstly, according to the sensitive degree of a salient pole characteristic and produced position vibration and vibration noise of a motor rotor, an injecting voltage amplitude and frequency suitable for a specific motor are chosen; then, according to determined parameters, pulse vibration high-frequency injection is carried out in an optimized injection angle scanning mode, and the peak position of a high-frequency current amplitude curve is extracted; at last, according to a nonlinear magnetization characteristic of a stator core, the N and S polarities of a permanent magnet are determined, and therefore the initial position of the rotor is determined. The surface-mounted type permanent magnet synchronous motor rotor initial position estimation method resolves various realistic problems of an initial position estimation stage, improves recognition precision, weakens negative influences of an injection process, and can effectively estimate any surface-mounted type permanent magnet synchronous motor rotor initial positions.

The invention relates to the communication technology field and discloses an ePDG selecting method and device, which can improve the ePDG selection accuracy without increasing the external signaling overhead and time delay when a terminal accesses an EPC network via an untrusted WLAN. The method comprises the steps of: receiving a domain name resolution request sent by a terminal, wherein the domain name resolution request carries an initial domain name and access network information of the terminal; searching for location information corresponding to the access network information in a correspondence table, wherein the correspondence table is used to indicate the corresponding relationship between the access network information and the location information; determining an IP address of an ePDG based on the initial domain name and the location information, transmitting the IP address of the ePDG to the terminal, and therefore conveniently selecting the ePDG by the terminal according to the IP address of the ePDG. The ePDG selecting method and device of the present invention are applicable to the process of selecting an ePDG.

The embodiment of the invention provides a video play method and device. The method comprises the steps: detecting whether the previewing operation of a target video identification in a preview video identification list is received or not when a terminal plays a preset video in a play window; determining a preview region on a screen of the terminal when the previewing operation is received; displaying a preview window in the preview region; and playing a video corresponding to the target video identification in the preview window. Through the method, the terminal can play the video corresponding to the target video identification selected by a use from the video identification list when the terminal plays the preset value, and determines whether the user is interested in the video or not according to the played video content. Compared with a mode that the user just selects the interesting video in the video list in the prior art according to the name of the video, the method enables the user to determine whether the user is interested in the video or not according to the content of the video, and can improve the selection accuracy.

The current invention is designed to improve the certainty of test-taker classifications in testing programs utilizing open-ended items. The current invention discloses methods to improve certainty in estimates of test-taker performance for all test-takers while using a minimum of scoring resources. The invention selectively allocates resources to scoring by ceasing to allocate scoring resources to test-takers whose performance has been determined with sufficient certainty, and by having an additional reader or readers score responses from test-takers whose scores are close to cutoff levels. The invention further increases precision of scores for test-takers near a cutoff by selectively allocating more reliable and accurate readers to score responses most likely to influence a test-taker's performance classification. Other scoring resources are also shown.

The embodiments of the invention provide a video playing method and device. According to the video playing method and device, when the playing window of a terminal plays a preset video, the terminal can also play a plurality of association videos associated with the preset video, and therefore, a user can watch the preset video and the association videos simultaneously; and when the user is not interested in the preset video played by the playing window, the user can select a video in which the user is interested according to the association videos which are played by a plurality of preset windows respectively. According to the prior art, a user selects one video episode according to the name or serial number of the video episode in a video list, and according to the video playing method and device provided by the embodiments of the invention, the user can select a video in which the user is interested according to the association videos which are played by the terminal, and compared with the prior art, the video playing method and device can improve selection accuracy.

According to the bolt type selection method, the bolt connection rigidity can be simulated more accurately, and the bolt transverse and axial loads given through finite element analysis are close to the actual situation. Bolt loads obtained through finite element analysis are used as input, then a mature and reliable empirical formula is used for checking and model selection, and bolt stress obtained through finite element analysis is not directly used for checking. The respective advantages of a finite element method and an empirical formula are fully utilized, and the model selection accuracy is improved. On the basis of the technical scheme, checking optimization of the bolt can be completed in the concept stage of vehicle model development. Reasonable model selection precision can be achieved only through finite element simulation and empirical formula calculation, no physical test is needed, and therefore the vehicle model development cost is saved, and the research and development period is shortened.

An embodiment of the invention provides a webpage topic sentence extraction method. The method comprises: firstly, obtaining a to-be-determined webpage, wherein the to-be-determined webpage contains a plurality of alternative topic sentences, and each alternative topic sentence contains a plurality of segmented words; secondly, determining a word feature value of each segmented word, inputting the word feature value into a preset machine learning model to obtain a partial order value of the segmented word, and further according to the partial order value of the segmented word, determining a partial order value of each alternative topic sentence; and finally, determining the alternative topic sentence with the partial order value greater than a preset threshold as a target topic sentence. According to the embodiment of the invention, the partial order values of the alternative topic sentences are obtained by utilizing the machine learning model; the machine learning model can reflect a degree of correlation between a query statement and a recalled webpage, so that the determined partial order values are more accurate and the accuracy of selecting the target topic sentence is improved. In addition, the invention furthermore provides a webpage topic sentence extraction apparatus, which is used for ensuring the application and implementation of the method in practice.

The invention relates to the technical field of calibration methods of line structured light vision systems, in particular to a laser line key point extraction method based on a plane target. The method is characterized by comprising the following steps: S1, binarization of a target image: converting a gray level image obtained by a CCD black and white camera into a binary image; S2, morphologicalimage processing: respectively carrying out corrosion and opening operation processing on the target binary image, and subtracting the result of the opening operation and the result of the corrosionoperation to remove useless information interference in the target image so as to obtain laser ray key image information; S3, Hough linear transformation: acquiring a laser line two-dimensional equation; S4, extracting of two-dimensional coordinates in the image coordinate system on the target laser line. The method has the advantages that the laser line equation extracted through the method can accurately calculate the image coordinate points on the laser line, and the measurement accuracy of the system is greatly improved.

The invention discloses a rice granule shape gene qSS7 as well as a preparation method and application. The sequence of the rice granule shape gene qSS7 is as shown in SEQIDNO.1. Due to the gene, the granule length and the length to width ratio can be increased, the granule width can be reduced, and the chalkiness rate and the chalkiness degree can be also reduced; and when the appearance quality of rice is improved by using the gene, the quality and the yield of rice can be prevented from negative influence; a gene mark is also designed for qSS7, long and short granule allele can be identified in the seedling period, unexpected heterozygous and pure short granule genotype plants can be rejected, the possibility that descendant granule shapes can be separated when being bred is avoided, and breeding procedures for a next season can be reduced. Therefore, the area of a descendant bred material can be reduced by 75% in quality breeding practice, the gene selection accuracy can be improved, and the breeding progress can be accelerated.

The invention belongs to the technical field of display, and specifically relates to a gray-scale voltage conversion method and module, a data driving circuit and a display panel. The gray-scale voltage conversion method is used for converting pixel data into display data including gray-scale information corresponding to the pixel data, wherein the pixel data is digital, and the display data is analog; splitting the pixel data into high byte data and low byte data; performing digital-analog conversion on the low byte data and the high byte data; and synthesizing the low byte data and the high byte data after the digital-analog conversion. The gray-scale voltage conversion method and module can reduce the selection delay time and improve the selection accuracy of the gray-scale voltage.

The invention discloses a multi-task and proximity information fusion depth learning algorithm based on object detection, which comprises the steps: pictures are input, image features by using a convolution neural network are extracted, and target candidate frames are generated; using the image features, the image is passed through the region candidate network to extract the target prediction frame; feature extraction and feature pooling of the target prediction frame, then the border regression, direction prediction, target detection and classification are performed to get the preliminary detection results; the preliminary detection result is fused with the target candidate frame and enters the ROI pooling layer and passes through the second full connection layer to obtain the final detection result; the classification of target detection is to redefine the confidence score of a target prediction box by using the information relationship between the target prediction box and its neighboring other target prediction boxes. The algorithm adopts multi-task output mode. The invention not only improves the speed of the target detection, but also ensures the accuracy of the target detection and achieves the requirement of the real-time target detection.

The invention discloses a node vitality quantifying method in a mobile social network, the method comprising: (1)configuring a connection strength index between mobile nodes; (2) computing a betweenness centrality of nodes; (3)counting, within a fixed time period, average access time rate of nodes and betweenness centrality, weighted-synthesizing them into a cooperation index of nodes; (4)computing and taking the minimum value of entropy of link information as link information entropy between node pairs; (5)taking an average value of entropy of link information between a mobile node and all reachable nodes in the network as a quantization value of the vitality of the mobile node. According to the invention, by introduction of several social factors as a supplement, a reasonable quantization is performed on the vitality of nodes in a mobile social network based on analysis of a mobile social network topology structure, and thus a scientific, reasonable and real result of mobile node vitality quantifying is provided for related applications. The invention improves node selecting accuracy rate in a mobile social network, and ensures reliable and high effective operation of mobile network service.

The invention provides a corn variety breeding method. The corn variety breeding method comprises the steps of drought resistance screening of a basic material, drought resistance screening of a selfing line F1-generation, drought resistance screening of a selfing line F2-generation, yield increase screening of a selfing line F3-generation, drought resistance screening of a selfing line F4-generation, hybrid preparation, drought-resisting and high-yield hybrid breeding and the like. In different generations, by adopting the method for identifying three types of drought resistance by introducing repeated drought-condition seedling survival rate, drought resisting indexes and 'three controls and three selections' according to the different characteristics of the generations, comprehensive assessment on drought resistance is performed, results are accurate and reliable, and the line selecting accuracy is improved.

The invention provides an SSR (Simple Sequence Repeat) marker tightly interlocked with a cucumber fruit bitter gene Bt and an application thereof, and relates to a biological breeding auxiliary technology. The sequence of the marker is as follows: SSR21558-F / SSR221558-R: GTGGGGGATGTGATT CAGAC / CATCATCCATTCCCCTCAAC, wherein the amplified characteristic stripe is (175bp) represented by Seq ID No.1 and (192bp) represented by Seq ID No.2; and SSR20054-F / SSR20054-R: GTTTGTGAGGGAAACGCAAT / TCAAAAAGCTTCCTTCCTTCA, wherein the amplified characteristic stripe is (127bp) represented by Seq ID No.3 and (108bp) represented by Seq ID No.4. Two markers are more tightly interlocked with the gene Bt so as to be more helpful to the establishment of a cucumber fruit bitter molecular marker breeding auxiliary system; with the adoption of the SSR marker obtained by the system, the screening of fruit bitterness or no fruit bitterness is carried out on a cucumber candidate material at any phase, so as to have the advantages of high efficiency, less limitation and accuracy, and improve the selection efficiency and accuracy for selecting bitter-free materials for cucumber breeding.

A method for optimizing design of building aseismic support and hanger type selection includes the steps of arranging and assembling main accessories in seismic support and hanger according to different type to obtain all combination types of main accessories of seismic support and hanger; calculating the design bearing capacity and cost of the aseismic support and hanger corresponding to each combination type are calculated by choosing the combination type which accords with the aseismic support and hanger layout form among all the combination types of the main accessories of the aseismic support and hanger. The design bearing capacity of each type of aseismic support and hanger is transformed into the maximum mass of the mechanical and electrical pipelines. According to the type and quality of mechanical and electrical pipelines and the maximum quality and cost of mechanical and electrical pipelines can be borne by the corresponding seismic support and hanger of each different combination type, the optimal combination type is selected in different combination types. The invention can effectively select the optimal combination scheme from the combination schemes of different typesof the main accessories of the seismic support and hanger, has high safety and low cost, and can meet the related requirements.

Embodiments of the invention disclose an information recommendation list generation method and device. The method comprises the following steps of: obtaining feature data of each to-be-recommended object in a to-be-recommended object set by a terminal, wherein the to-be-recommended object set comprises S1 to-be-recommended objects; obtaining a preset feature data set by the terminal, wherein the preset feature data set comprises feature data of S2 recommended objects, the feature data of the S2 recommended objects comprises feature data of an appointed recommended object, and S2 is smaller than or equal to S1; calculating a recommendation value of each to-be-recommended object by the terminal according to the preset feature data set and the feature data of the S1 to-be-recommended objects;and selecting a target recommended object by the terminal according to the recommendation value of each to-be-recommended object in the S1 to-be-recommended objects, and adding the target recommendedobject to an appointed display position in a recommendation list. By adoption of the method and device, the advantages of improving the recommended object selection correctness and improving the resource utilization rate of the recommendation list are provided.

The invention discloses a sub-block motion prediction and coding method, a coder and a storage device, and the method comprises the steps: obtaining a reference block, a co-located frame and a reference frame of a current block, and the number of at least one of the reference block, the co-located frame and the reference frame being greater than one; obtaining motion vectors of the current blocks,wherein the motion vector of each current block corresponds to a group of reference parameters; obtaining a common position block of a plurality of sub-blocks based on each group of reference parameters, the plurality of sub-blocks being obtained by dividing the current block, the common position block of the sub-blocks being a block where a first offset point is located in a co-located frame, and the first offset point being a point after a specified point in the sub-blocks offsets a reference motion vector; obtaining candidate motion vectors of the sub-blocks by utilizing the motion vectorsof the common position blocks of the sub-blocks; and obtaining a sub-block motion prediction result of the current block based on all candidate motion vectors of the sub-block according to the principle of minimum prediction cost. In this way, the prediction accuracy can be improved.

A nerve stimulation apparatus is provided. The nerve stimulation apparatus detects nerve activities from a body surface and applies stimulation. The nerve stimulation apparatus includes a stimulation apparatus with multiple electrodes which are arranged on skin and a current supply unit which supplies a current to the electrodes, which stimulation apparatus provides the current to a living body percutaneously to stimulate a target nerve; a measurement apparatus which measures activities of muscles governed by the nerve according to the stimulation from the stimulation apparatus; and an information processing apparatus which determines, based on a measurement result of the nerve activities obtained from the measurement apparatus, which electrode is capable of providing the target nerve activities equal to or greater than a desired value.