34results about How to "Data quality relief" patented technology

A system and method for actively selecting and labeling images for semantic segmentation are disclosed. A particular embodiment includes: receiving image data from an image generating device; performing semantic segmentation or other object detection on the received image data to identify and label objects in the image data and produce semantic label image data; determining the quality of the semantic label image data based on prediction probabilities associated with regions or portions of the image; and identifying a region or portion of the image for manual labeling if an associated prediction probability is below a pre-determined threshold.

In order that it is realized to put 16× speed DVD-RAM in practice, a recording method coping with CAV (constant angular velocity) and a method for improving S / N ratio are provided and an optical disk device using them is provided. In order that a cutoff phenomenon of pulse in a laser driving circuit may not occur, the shortest pulse width is set at 1 Tw or more where Tw indicates a window width, and a recording pulse in which a 4T mark is recorded with a mono pulse where T indicates a detection window width and the parameter setting range are prescribed. Simultaneously, a reproduction power for 6× speed or more is set at 1.5-2.0 mW, different from a conventional value, and a setup sequence with which deterioration of data in reproducing does not occur is provided. Thereby it becomes possible to put 16× speed DVD-RAM in practice.

Improved sensors and sensing methods for detection of Neuromuscular Blockade (NMB), for example to improve monitoring generally, as well as facilitate automated NMB drug administration. The methods, systems, devices, etc., herein can increase the quantity and quality of data available. The methods, etc., include translation between commonly available sensing types, use of partially saturated sensor measurements, use of a model of neurotransmitter storage and release and sophisticated use of more than one sensor simultaneously. These methods are aided by new sensors and methods of decreasing electrical noise. An implementation of the methods, systems, devices, etc., herein is automated drug delivery for NMB, whether through an advisor system or through with full computer control.

Discrete Cosine Transforms in a radix-2 block transform method enables true split and merge transform processing of equal sized data halves in the transform domain for enabling recursive merges and splits in transform domain without data degradation. Input data in the time domain or spatial domain during either the split and merge radix-2 forward transform processing is firstly transform processed followed by combinational processing. In the split transform process, whole transformed data is split using combinational processing into first and second transformed data halves in the transform domain as a true split. In the merge transform process, first and second transform halves are merged using combinational processing into a merged transformed data whole in the transform domain. In either case, time or spatial domain input data can be transformed into the transform domain in the form of split halves or merged whole. The split halves can be merged by the merge process combinational processing and the merged whole can be split by the split process combinational processing. The split and merge process combinational processing enable recursive splits and merges in the transform domain.

Methods for non-specifically amplifying a nucleic acid template molecule are provided. The methods may be used to amplify nucleic acid template molecule(s) for sequencing, e.g., for sequencing the genomes of uncultivable microbes or sequencing to identify copy number variation in cancer cells. Aspects of the disclosed methods may include non-specifically amplifying a nucleic acid template molecule, including encapsulating in a microdroplet a nucleic acid template molecule obtained from a biological sample, introducing multiple displacement amplification (MDA) reagents and a plurality of MDA primers into the microdroplet, and incubating the microdroplet under conditions effective for the production of MDA amplification products, wherein the incubating is effective to produce MDA amplification products from the nucleic acid template molecule.

In this invention, wireless video telephone communication between mobile telephones MS1 and MS2 using the MPEG-2 standard is performed when the reception field strength falls below a threshold value of H continuously for a specific period of T3 or longer in the course of receiving the data at one mobile telephone MS2, the telephone MS2 monitors the restoration of the reception field strength. After detecting the restoration, mobile telephone MS2 transmits a request for the transmission of an I frame to mobile telephone MS1. Upon receipt of the request, mobile telephone MSI creates an I frame in place of a P frame and transmits, the I frame.

There is provided a communication control apparatus including an acquisition section configured to acquire first information related to communication quality of a wireless channel through which target data to be distributed to a terminal apparatus is transmitted to the terminal apparatus, and second information related to data among the target data, the data being accumulated in the terminal apparatus, and a control section configured to perform control over transmission of the target data to the terminal apparatus on the basis of the first information and the second information.

The present disclosure relates to new nucleotide and oligonucleotide compounds and their use in nucleic acid sequencing applications.

A method is indicated for determining a confidence value of an object of a class detected in an input image with the aid of a trained neural network, including: producing an activation signature for the class of the detected object using a plurality of output images of a layer of the neural network, the input image being provided to the input of the neural network; scaling the activation signature to the dimension of the input image; comparing an object portion of the scaled activation signature with an activation signature distribution of all objects of the same class of a training data set of the neural network in order to determine the confidence value.

Methods and systems are provided for aggregating features in multiple video frames to enhance tissue abnormality detection algorithms, wherein a first detection algorithm identifies an abnormality and aggregates adjacent video frames to create a more complete image for analysis by an artificial intelligence detection algorithm, the aggregation occurring in real time as the medical procedure is being performed.

The invention relates to a system and a method for detecting the air gap of a wind power generator in real time. The system comprises a wind power generator, a laser sensor and a capacitive sensor, wherein the distance between the stator silicon steel sheet of the wind power generator and the rotor magnetic steel is an air gap of the wind power generator, at least one laser sensor is fixedly connected to the rotor shell through a mounting support; at least one capacitive sensor is pasted on the surface of the stator silicon steel sheet, and a laser beam emitted by the laser sensor is irradiated on the surface of the capacitive sensor, a set of sensor measuring devices are formed for measuring the air gaps of the wind power generator. According to the system and the method, the load of a fan can be evaluated in real time, the load can be controlled in advance through the fan, and the stator and the rotor can be prevented from being scratched.

Fast Hartley Transforms in a radix-2 block transform method enables true split and merge transform processing of equal sized data halves for enabling recursive merges and splits in Hartley transform domain without data degradation. Input data in the time domain or spatial domain during either the split and merge radix-2 forward Hartley transform processing is firstly transform processed followed by combinational processing. Whole transformed data is split using combinational processing into transformed data halves in the Hartley transform domain as a true split. The transformed halves are merged using combinational processing into whole merged transformed data in the Hartley transform domain. Time or spatial domain input data can be transformed into the Hartley transform domain in the form of split halves or merged whole. The split halves can be merged by the merge process combinational processing and the merged whole can be split by the split process combinational processing.

The invention discloses a generalized RQD acquisition method for rock mass which eliminates the influence of measuring blank space, comprising: adopting a photogrammetric method for the rock mass to be treated to establish a three-dimensional model of the rock mass slope; adopting a space disk rule according to the measured data of the model to establish a three-dimensional model of the discontinuous surface of the rock mass; acquiring information of the blank space; setting at least one filler for filling the blank space, obtaining a three-dimensional model of the discontinuous surface havingat least one filler; for the three-dimensional model of the discontinuous surface having at least one filler, selecting the optimal discontinuous surface spacing threshold corresponding to the X-axis, Y-axis and Z-axis direction line groups by traversing method, and obtaining the generalized RQD of the rock mass eliminating the influence of blank space according to the spatial relationship between the lines, and the space disc and the fillers. The invention obtains the generalized RQD of the rock mass eliminating the influence of blank space by the measured data of the discontinuous surface,and may more accurately evaluate the integrity quality of the rock mass.

Astigmatism and coma are suppressed in an optical pickup device such that light from a single real laser light source is diffracted and divided by a hologram module into a plurality of light spots which are focused upon a plurality of tracks of an optical disk to read data on the tracks at the same time. A non-diffraction end of fraction hologram patterns of the hologram module are determined so that light diffracted by the hologram module is given inverse aberration of that to be caused by optical elements in optical path from the real laser light source to the optical disk. Another hologram pattern provides a uniform intensity of a light spot formed of a laser light by an optical element.

The invention provides a multi-source client data processing method and device, a platform and a storage medium, which relate to the technical field of multi-source data processing. The method comprises the steps of obtaining a basic data table of each data source in a plurality of data sources, wherein the basic data table of each data source stores original data of at least one client attributefrom each data source, generating an integrated data table according to the basic data tables of the plurality of data sources, the integrated data table storing at least one client attribute, an association relationship between the at least one client attribute and the original attribute group, and a priority of the at least one client attribute, and an original attribute field corresponding to each client attribute in the basic data tables of the plurality of data sources being stored in the original attribute group associated with each client attribute, and the priority of at least one client attribute being the priority of each data source corresponding to at least one client attribute. By means of the scheme, the CDP platform is ensured to have the most complete client data, and client data loss is avoided.

The invention discloses a distribution network operation inspection-oriented multi-source data acquisition method, which adopts a micro-service framework, provides data information needing to be synchronized through a target system, and comprises the following steps of: S1, actively or passively pushing the data information by a data source system; S2, after the target system receives the request, performing the following steps: S21, for different data receiving modes, respectively providing different analysis data models, and performing data analysis; S22, performing data caching; S23, processing the queue cache by adopting an asynchronous multithread; S24, establishing a mapping relationship between the synchronization field and the database field, keeping the field attributes consistent, and performing data mapping; and S25, presisting the data into a database table, and completing data synchronization. According to the invention, through establishment of the collection specification of the multi-source data, the access flexibility of the multi-source data under multiple platforms is improved, the data quality problem is reduced, the system application capability is improved, the inspection system application capability is improved, the equipment research and judgment accuracy is ensured, and the reliability of the net rack is improved.

The invention provides a supervised learning-based house type layout generation method and system, and the method comprises the steps: building a house type layout generation model which comprises a generation network, a simulation environment and a scoring network; acquiring house type example data, inputting the house type example data into the generative network, and outputting house type layout data; according to the house type example data, house type design conditions are obtained, the house type design conditions and the house type layout data are input into a simulation environment, and the house type layout is scored; according to the evaluation network, taking the house type layout data as input data, taking the score output by the simulation environment as a mark, training the evaluation network and outputting score data; back-propagating the score data to the generative network, updating the parameters of the generative network, training the generative network, and stopping training when the output result of the generative network meets a preset requirement to obtain a trained house layout generation model; and obtaining house type design conditions input by a user, inputting the house type design conditions to the trained house type layout generation model, and outputting a house type layout map.

The invention provides a suspended microstrip filtering device and a microstrip filtering device. The suspended microstrip filtering device comprises a PCB which comprises a substrate including at least one resonator, a grounding board, and an insulated separator which is inserted between the PCB and the grounding board and includes multiple inflation openings.

A generalized radix-2 block transform method enables true split and merge transform processing of equal sized data halves in the transform domain enabling recursive merges and splits in transform domain data degradation. Input data in the time domain or spatial domain during either the split and merge radix-2 forward transform processing is firstly transform processed followed by combinational processing. Whole transformed data is split using combinational processing into transformed data halves in the transform domain as a true split. The transformed halves are merged using combinational processing into whole merged transformed data in the transform domain. Time or spatial domain input data can be transformed into the transform domain in the form of split halves or merged whole. The split halves can be merged by the merge process combinational processing and the merged whole can be split by the split process combinational processing.

The invention discloses a minimum closed loop automatic forming method applied to disconnected pattern spots, which comprises the following steps of setting a buffer distance, finding all required pattern spots according to a pattern topological relation among the pattern spots, and automatically forming a minimum closed loop based on the found pattern spots. According to the method provided by the invention, existing people engaged in the work are liberated from complicated data, so that the data quality problem caused by misoperation of operators is reduced, the data processing efficiency and quality are improved, and the data processing cost is saved.

The invention provides a multi-bank participated anti-money laundering discrimination method and related application equipment, which can be applied to the financial field, and the method comprises the steps: inputting obtained money access operation information of a target customer into a central privacy calculation data model, and obtaining a money laundering risk probability factor of the target customer; the central privacy computing data model is obtained by fusing and optimizing models uploaded by various bank databases; the model uploaded by each bank database is obtained by training the corresponding central privacy computing data model by using own data; when it is judged that the money laundering risk probability factor exceeds a preset threshold value, it is judged that the target customer has a money laundering operation risk; that is, according to the method, barriers among multiple bank systems can be broken, the privacy computing technology is utilized, encryption parameters among banks are exchanged, joint computing modeling is carried out, and the problems that according to an existing related scheme, whether a customer has a risk behavior or not can be judged only through data of a single bank, and the judgment precision is low are solved.

Fast Fourier Transforms in a radix-2 block transform method enables true split and merge transform processing of equal sized data halves in the Fourier transform domain for enabling recursive merges and splits in Fourier transform domain without data degradation. Input data in the time domain or spatial domain during either the split and merge radix-2 forward transform processing is firstly transform processed followed by combinational processing. Whole transformed data is split using combinational processing into transformed data halves in the transform domain as a true split. Transform halves are merged using combinational processing into whole merged transformed data in the transform domain. Time or spatial domain input data can be transformed into the transform domain in the form of split halves or merged whole. The split halves can be merged by the merge process combinational processing and the merged whole are split by the split process combinational processing.

Discrete Sine Transforms in a radix-2 block transform method enables true split and merge transform processing of equal sized data halves in the DST transform domain for enabling recursive merges and splits in DST transform domain without data degradation. Input data in the time domain or spatial domain during either the split and merge radix-2 forward transform processing is firstly transform processed followed by combinational processing. Whole transformed data is split using combinational processing into transformed data halves in the transform domain as a true split. The transformed halves are merged using combinational processing into whole merged transformed data in the transform domain. Time or spatial domain input data can be transformed into the transform domain in the form of split halves or merged whole. The split halves are merged by the merge process combinational processing and the merged whole are split by the split process combinational processing.

The invention discloses a GPS equipment installation quality verification method and device, equipment and a storage medium, relates to the technical field of data verification, and can solve the technical problems of low verification precision and high cost caused by single verification data dimension and relatively high data specificity during GPS equipment installation quality verification. Comprising the following steps: acquiring a construction contract uploaded by a GPS equipment data user, and extracting construction work order information corresponding to GPS equipment installation according to the construction contract; construction result information of a construction party after construction according to the construction work order information is determined, and verification item types included in the construction result information and the construction work order information are the same; and comparing the construction work order information with the construction result information to determine an installation quality verification result of the GPS equipment. The method is suitable for realizing intelligent verification of the installation quality of the GPS equipment from a plurality of data dimensions contained in the construction work order information and the construction result information.

The invention provides a suspended microstrip filter device and a microstrip filter device. A suspended microstrip filter device includes a printed circuit board including a substrate having at least one resonator thereon; a ground plate; and an insulating separator interposed between the printed circuit board and the ground plate, the The insulating separator has a plurality of openings filled with air.