43 results about "Feature relevance" patented technology

Membrane proteins are difficult to express in recombinant form, purify, and characterize, at least in part due to their hydrophobic or partially hydrophobic properties. The membrane scaffold proteins (MSP) of the present invention assemble with target membrane or other hydrophobic or partially hydrophobic proteins or membrane fragments to form soluble nanoscale particles which preserve their native structure and function; they are improved over liposomes and detergent micelles. In the presence of phospholipid, MSPs form nanoscopic phospholipid bilayer disks, with the MSP stabilizing the particle at the perimeter of the bilayer domain. The particle bilayer structure allows manipulation of incorporated proteins in solution or on solid supports, including for use with such surface-sensitive techniques as scanning probe microscopy or surface plasmon resonance. The nanoscale particles, which are robust in terms of integrity and maintenance of biological activity of incorporated proteins, facilitate pharmaceutical and biological research, structure/function correlation, structure determination, bioseparation, and drug discovery.

The invention solves the problems such as feature extraction, asynchronization and synchronization, feature relevance and data difference value to fuse visual data and laser radar data so as to generate three-dimensional color depth image data, namely a three-dimensional color laser data point cloud, thereby providing precise distance information and rich environment information, effectively overcoming the defects of the traditional laser radar and the visual sensor in the aspects of robustness, information amount, precision and cost. The invention integrates a two-dimensional laser radar, a monocular vision sensor, an inertial navigation data measuring unit, a driving motor and the like into an information fusion-based three-dimensional color laser radar integral unit with universal applicability and high precision, and utilizes inertial navigation data and visual data to perform secondary alignment on laser radar data, the structure is simple, equipment does not need to be subjected to precise calibration, the performance is reliable and stable, the practicability is strong, and the method and the device can be widely used in various indoor and outdoor environments under the motion condition, and have certain universal applicability.

The invention provides a wireless terminal and a remote parameter setting method thereof. When a dial-up connection is established by the wireless terminal, a request message to update parameters is sent to a network server by the wireless terminal; equipment information in the request information of the wireless terminal matches with equipment information of a server through the network server; parameter setting information which matches with the equipment information of the wireless terminal is sent; and the parameters are set by using the received parameter setting information through the wireless terminal. The users not only enjoy new function experience brought by technical development, but also do not worry about inconvenience caused by that a wireless router cannot be normally used due to being not familiar with new functions, and further do not worry about complaints from customers or user group churn, which is caused by that the users can not use the terminal for a short time because of updating; and meanwhile, a remote setting function can be simply implanted to all platforms of a data card due to having low relativity with other functions. Furthermore, the remote parameter setting method is simple in operation, and the users can enjoy convenient services.

Membrane proteins are difficult to express in recombinant form, purify, and characterize, at least in part due to their hydrophobic or partially hydrophobic properties. Membrane scaffold proteins (MSP) assemble with target membrane or other hydrophobic or partially hydrophobic proteins or membrane fragments to form soluble nanoscale particles which preserve their native structure and function; they are improved over liposomes and detergent micelles. In the presence of phospholipid, MSPs form nanoscopic phospholipid bilayer disks, with the MSP stabilizing the particle at the perimeter of the bilayer domain. The particle bilayer structure allows manipulation of incorporated proteins in solution or on solid supports, including for use with such surface-sensitive techniques as scanning probe microscopy or surface plasmon resonance. The nanoscale particles facilitate pharmaceutical and biological research, structure/function correlation, structure determination, bioseparation, and drug discovery.

A device (100) for managing a plurality of software items, the device (100) comprising an analysis unit (103) adapted for analyzing a functional correlation between the plurality of software items, and a grouping unit (105) adapted for grouping functionally correlated ones of the plurality of software items together in a common memory space.

The invention relates to an image super-resolution method and system based on depth feature relevance. The method comprises the following steps: acquiring a low-resolution image to be reconstructed; inputting the low-resolution image to be reconstructed into a trained student model for reconstruction to obtain a high-resolution image, wherein the training process of the student model comprises the following steps: taking a low-resolution training image as the input of the student model; taking the high-resolution training image as the output of the student model, taking the overall loss function as a loss function, and utilizing a loss function attenuation mechanism to supervise and train the student model through the high-resolution image and the real image output by the trained teacher model to obtain the trained student model. According to the invention, the high-resolution image is estimated through the image super-resolution model which is small in parameter quantity and high in performance.

The invention discloses a terminal and a terminal function option display method. The terminal comprises a function module, a function option module, a classification module and a display module, wherein the function module is used for providing different terminal functions for a user; the function option module is used for generating different function options and sub-function options according to the different terminal functions of the function module; the classification module is used for classifying the different function options and sub-function options into different tree-shaped branches according to the function dependence of the different function options and sub-function options, such as trunks, branches and leaves, and is used for representing the different function options and sub-function options of every executable function of the terminal; and the display module is used for displaying all function options and sub-function options classified by the classification module in a tree shape, such as the trunks, the branches and the leaves. The terminal function option display method can bring good impression to users and bring convenience to user operation.

The invention discloses a feature extraction method, device and server, and a computer readable memory medium and belongs to the field of the mobile Internet. The method comprises the steps of: pre-grouping a pre-training data set of a wireless side and a user side according to feature relevance, thereby obtaining initial model parameters; carrying out grouping feature relevance computing on the training data set according to the initial model parameters, classifying features of the training data set into corresponding groups, and updating the model parameters; and carrying out iterative optimization on feature contraction variables of each group of features, thereby obtaining a feature selection result of the wireless side and the user side. According to the method, the device, the serverand the computer readable memory medium, through utilization of a feature grouping selection mechanism, feature samples are pre-grouped before feature extraction is carried out; the grouping featurerelevance computing can be carried out on newly increased features; the feature redundancy computing problem can be effectively solved; grouping optimization is further carried out on grouping features through importing of the feature contraction variables; feature grouping selection efficiency is greatly improved; and online dynamic data feature extraction can be supported.

The invention relates to an attention-fused single image rain removal method, which comprises the following steps of: firstly, carrying out batch normalization processing on an input image, and then inputting the input image into an encoder network in which a three-dimensional attention mechanism and a residual dense block structure are combined to obtain high-dimensional features; the global feature relevance is calculated by using a Transform mechanism; reducing the feature matrix to the size of an original input image step by step by using a decoder constructed by a three-dimensional attention residual dense block to obtain an output image; in order to solve the problem that image structure information and detail high-frequency information are erased due to an algorithm in the rain removal process, multi-scale structural loss and a common rain removal loss function are combined to participate in rain removal network training. According to the invention, a rain-removed image with higher quality can be obtained by carrying out rain removal on a single rain image. According to the method, the three-dimensional attention mechanism, the Transform and the coder-decoder architecture are combined, so that the rain removal performance of the network can be better improved.

A system for feature relevance visualization optimization is provided. The system comprises a controller configured for modifying placement of features in a relevance visualization. The controller is further configured to: receive interaction data comprising one or more features positioned in the relevance visualization, wherein the one or more features are defined and measurable properties of the interaction data; construct a logical grouping of the one or more features based on a type of each of the one or more features, wherein similar features are collocated in the relevance visualization; construct a machine learning-based grouping of the one or more features based on relevance calculations for the one or more features; combine the logical grouping and the machine learning-based grouping to generate a combined feature placement, wherein the one or more features are repositioned in the relevance visualization; and output the relevance visualization having the combined feature placement.

The invention provides a multi-modal 3D target detection method and a system thereof. The method comprises the following steps: respectively extracting features of an original image I and a corresponding LiDAR point cloud L; carrying out point and pixel feature fusion on the features of the original image I and the corresponding LiDAR point cloud L to form LiDAR point cloud features, taking the features of the original image I as image features, and respectively generating a 3D region proposal and a 2D region proposal; extracting features from the 3D area proposal and the 2D area proposal respectively and fusing, and generating a final 3D target detection result. Meanwhile, the invention provides a corresponding terminal and a medium. According to the method, target detection is completed by using a geometric constraint relationship and feature relevance between modals; a 3D target detection task is completed through feature fusion of a point-pixel level of a first stage and a region proposal level of a second stage; and a high-quality regional proposal is generated by using geometric constraints between the image and the LiDAR point cloud.

The invention provides a screening method of individualized medicine. The method includes: building a disease-related pathogenic protein molecule database, building a paint individual clinical disease course symptom-protein pathogenic factor response feature relevance database, building a disease model target medicine-protein response relevance database, and acquiring a medicine screen result. The screening method has the advantages that the method is brand new and pioneering and used for individualized precise medical treatment, genetic and environment pathogenic factors and patient disease course features can be accurately and comprehensively reflected in the detecting and analyzing result of protein molecules, and objective, precise and individualized treatment medicine applicable to different patient disease course features is screened according to the target medicine's correction of the pathogenic protein molecules.