799 results about "Branch structure" patented technology

A memory module includes at least one CAR and a plurality of DRAMs provided so as to be close and adjacent to one another on one face and the other face of a module substrate. The DRAMs are divided into a plurality of memory groups. Memory groups adjacent to each other of these memory groups are paired with each other. One of this pair is a 1-ranked memory group and the other is a 2-ranked memory group. This pair of the memory groups is connected to the CAR via short wiring with a T-branch structure having a short stub. One of the pair of the memory groups on the signal-reception side functions as an open end. Active termination is performed by a termination resistor of the other of the pair of the memory groups on the signal-non-reception side. Subsequently, signal reflections can be reduced.

The embodiment of the invention provides a neural network training and construction method and device for object detection, and an object detection method and device based on a neural network. The neural network training and construction method for object detection comprises the steps: inputting a training image comprising a training object into the neural network, so as to obtain a prediction boundary frame of the training object; obtaining a first loss function according to the ratio of the intersection area and union area of the prediction boundary frame and a real boundary frame, wherein the real boundary frame is the boundary frame of the training object marked in the training image in advance; and adjusting the parameters of the neural network at least through the first loss function, so as to carry out the training of the neural network. According to the invention, the first loss function is employed for the regression of a target boundary frame into an integrated unit, and the object detection precision of the neural network is remarkably improved. Moreover, the training and detection efficiency of the neural network can be effectively improved through two branch structures of the neural network.

A 4-freedom parallel robot structure with passive constraint branch is disclosed, which contains different two passive constraint branch structures and can be used for robot to realize one axial movement and three different rotations. Its advantages are high parallel rigidity and load-bearing power, simple structure, high precision and low cost.

To provide the heat-meltable fluorine-containing resin fiber having excellent intermingling property and various materials for fiber products obtained therefrom. The heat-meltable fluorine-containing resin such as ETFE having a branched structure is used solely or mixed to other fiber such as an electrically conductive fiber to be formed into a non-woven fabric.

The invention discloses a multi-modal action recognition method based on a depth neural network. The method comprehensively utilizes multi-modal information such as video images, optical flow diagramsand human skeleton. The specific steps are as follows: firstly, a series of preprocessing and compression are performed on the video; Obtaining an optical flow graph based on adjacent frames of the video; The human skeleton is obtained from the video frame by using the attitude estimation algorithm, and the path integral characteristics of the skeleton sequence are calculated. The obtained optical flow graph, skeleton path integration feature and original video image are input into a depth neural network with multi-branch structure to learn the abstract spatio-temporal representation of humanmotion and correctly judge its motion category. In addition, the pooling layer based on attention mechanism is connected to the video image branch, which enhances the abstract features closely related to the final motion classification results and reduces the irrelevant interference. The invention comprehensively utilizes multimodal information, and has the advantages of strong robustness and high recognition rate.

The invention discloses a real-time robust far infrared vehicle-mounted pedestrian detection method. The method comprises the steps of catching a potential pedestrian pre-selection area in an input image through a pixel gradient vertical projection, searching an interest area in the pedestrian pre-selection area through a local threshold method and morphological post-processing techniques, extracting a multi-stage entropy weighing gradient direction histogram for feature description of the interest area, inputting the histogram to a support vector machine pedestrian classifier for online judgment of the interest area, achieving pedestrian detection through multi-frame verification and screening of judgment results of the pedestrian classifier, dividing training sample space according to sample height distribution, building a classification frame of a three-branch structure, and collecting difficult samples and a training pedestrian classifier in an iteration mode with combination of a bootstrap method and an advanced termination method. According to the real-time robust far infrared vehicle-mounted pedestrian detection method, not only is accuracy of pedestrian detection improved, but also a false alarm rate is reduced, input image processing speed and generalization capacity of the classifier are improved, and provided is an effective night vehicle-mounted pedestrian-assisted early warning method.

An image acquisition unit acquires a real endoscope image. A first evaluation value acquisition unit acquires a first evaluation value indicating the hole likeness of the insertion destination of an endoscope, and a second evaluation value acquisition unit acquires a second evaluation value indicating the boundary likeness of a plurality of bronchi at the bronchial branch. A determination unit determines whether or not a branching structure is included in the real endoscope image using both the first and second evaluation values. When it is determined that a branching structure is included, a virtual endoscope image generation unit generates a virtual endoscope image, and a display control unit displays the real endoscope image and the virtual endoscope image on a display.

A connecting branch structure for the parallel robot is composed of the sequentially connected multi-freedom hinge, parallelogram mechanism and moving set. A six-freedom parallel robot using said connecting branch structure consisting of movable platform, fixed platform and three said connecting branch structures. Its advantages are high flexibility and high rigidity.

The invention provides a new phosphorous acid concrete super plasticizer with a polyethyleneimine structure as well as a preparation method and an application thereof. The phosphorous acid concrete super plasticizer with a polyethyleneimine structure can effectively improve the concrete fluidity, and the preparation method is simple and effective. The preparation method comprises the following steps: a) making chloro-substituted polyether react with polyethyleneimine so that a part of amino groups H of the polyethyleneimine are substituted to obtain polyethyleneimine with a polyoxyethylene ether branch structure, wherein the molar ratio of the chloro-substituted polyether to the imino groups contained in the polyethyleneimine is 1:(3-8); and b) performing phosphorylation of the remaining amino groups H of the polyethyleneimine with a polyoxyethylene ether branch structure to obtain the phosphorous acid concrete super plasticizer with a polyethyleneimine structure.

The invention discloses a de-emulsifying agent and a method for preparing the same. The de-emulsifying agent is a polyether compound prepared by alkoxylation reaction of poly (amidoamine) (PAMAM) and epoxy compounds. The invention adopts the PAMAM with a dendritic multi-branch structure as an initiator for synthesis the de-emulsifying agent. The obtained de-emulsifying agent has the characteristics of a plurality of branched and large molecular weight. The de-emulsifying agent has good interfacial activity, infiltrability, good crude oil emulsion de-emulsifying performance, and wide applications in crude oil exploration, oil refining and other fields.

The invention discloses a steel rail tread defect recognition method based on combination of a gray image and a depth image. The method comprises the steps of establishing a steel rail tread image data set by utilizing a registered gray image and depth image pair of a steel rail tread, dividing the data set into a training sample set and a test sample set, and preprocessing the images of the dataset; performing defect recognition by adopting a special convolutional neural network structure, wherein a front end of the network is provided with two branch structures, which can extract features from the gray image and the depth image of the steel rail tread respectively; fusing feature information of the gray image and the depth image through feature graph connection, and then outputting preliminary prediction results by adopting a prediction module; and finally, screening the preliminary defect prediction results by adopting a non-maximum suppression method to obtain a final steel rail defect recognition result. The method combines two-dimensional and three-dimensional features of the steel rail tread at the same time, has the capability of distinguishing real defects and pseudo-defects, can reduce the misjudgment rate and the missing detection rate, and is especially suitable for defect recognition in a complex environment.

When a network pages the temporary user mobile identifier of a mobile station, the mobile station sends a response to the network. Next, the network checks the authenticity of the user using a ciphering key, corresponding to the temporary user mobile identifier and a random number. If the temporary user mobile identifier is authenticated, a normal incoming call acceptance procedure is executed. If the mobile station is authenticated although the temporary user mobile identifier is wrong, the network reassigns a new temporary user mobile identifier to the mobile station and stops the current communication. In communication, the network and the mobile station mutually notify encipherment-onset time and negotiate about encipherment manner with each other. In addition, diversity handover is commenced upon a call attempt. Furthermore, if a branch replacement is necessary, the current branch is replaced by new branches capable of executing the diversity handover. Additionally, when a new call occurs to or from the mobile station capable of treating a plurality of calls simultaneously, the mobile station uses the same branch structure and the same communication frequency band for all of calls. Additionally, when a new call occurs to or from the mobile station capable of treating a plurality of calls simultaneously, a branch structure and a communication frequency band, which can continue all of the calls, are selected and used. Therefore, the mobile communications system is suitable for transmission of various sorts of data in accordance with the development of multimedia.

Arterial and venous endothelial cells are molecularly distinct from the earliest stages of angiogenesis. This distinction is revealed by expression on arterial cells of a transmembrane ligand, called EphrinB2 whose receptor EphB4 is expressed on venous cells. Targeted disruption of the EphrinB2 gene prevents the remodeling of veins from a capillary plexus into properly branched structures. Moreover, it also disrupts the remodeling of arteries, suggesting that reciprocal interactions between pre-specified arterial and venous endothelial cells are necessary for angiogenesis. This distinction can be used to advantage in methods to alter angiogenesis, methods to assess the effect of drugs on artery cells and vein cells, and methods to identify and isolate artery cells and vein cells, for example.

The invention relates to a high fused mass strength polypropylene contained long-chain branch and the preparation method thereof. The high fused mass strength polypropylene resin is prepared by polypropylene grafted object, amine or alcohols compound as well as antioxidant and heat stabilizer, etc. The invention adopts a melting and continuously extrusion method with a reaction-type double screw extruder to prepare the high fused mass strength polypropylene, and uses supercritical carbon dioxide to assist the extrusion, which can lower the extrusion temperature, reduce the degradation of polymer and meanwhile greatly strengthen the pervasion of molecule, so as to further promote the reaction. In the reaction process, as chain extender, amine and alcohols compound react with special functional group of polypropylene grafted object, link linearity polypropylene molecular chain, consequently produce novel polypropylene that each grafted point includes two long-chain branch structures and of which the melt flow rate is adjustable. The novel polypropylene is characterized by non-gelatin and strain hardening. The product shows great advantages over a plurality of aspects like color, mechanical property and workability, therefore, the product is suitable for application areas of blowing, thermo-forming, blown film, extrusion and coating, etc.

Extracting a branch structure and a surrounding structure area in which a surrounding structure of the branch structure exists from a three-dimensional medical image, setting an attention branch structure in the branch structure, estimating an attention surrounding structure area functionally associated with the attention branch structure from the surrounding structure area based on the branch structure, obtaining an evaluation value in the attention surrounding structure area, and generating a mapped image which is an image generated by mapping the evaluation value in the attention surrounding structure area to a morphological image representing morphology of at least one of the attention branch structure and the attention surrounding structure area.

The invention discloses a cross-scale biomimetic micro-nano branch structure array and a preparation method thereof, which relate to the field of micro-nano materials and robots, particularly relate to a biomimetic adhesive material and a preparation method thereof and also relate to an electrochemical anodic oxidation process. The cross-scale biomimetic micro-nano branch structure array consists of a substrate and a micro-nano multilevel branch structure array, wherein the micro-nano multilevel branch structure array at least comprises a primary micron cylindrical structure array and a branch micro-nano cylindrical structure array. The preparation method thereof comprises the following steps of: firstly, etching an aluminum template with an argon ion beam to obtain an aluminum template with a primary micron hole array; secondly, preparing a micro-nano branch structure array in a definite point and direction on the template array by an electrochemical anodic oxidation process; thirdly, performing film extrusion and curing on the template by a polymer replica method; and finally, removing the template for forming to obtain the desired micro-nano structure array to form a biomimetic dry adhesive material. The cross-scale biomimetic micro-nano branch structure array and the preparation method thereof use electric characteristics to effectively solve the control problem of from micron to nanometer cross-scale branch structure arrays and have a great significance for the development of biomimetic adhesive materials and related dry biomimetic robots.

Heat exchange device with a boiling surface comprising porous surface layer arranged on a solid substrate, the porous surface layer comprises a porous wall structure defining and separating macro-pores that are interconnected in the general direction normal to the surface of the substrate and have a diameter greater than 5 μm and less than 1000 μm wherein the diameter of the pores gradually increases with distance from the substrate wherein the porous wall structure is a continuous branched structure.

The present invention is to provide oily makeup cosmetics which comprises an oligo ester obtained from (a) an aliphatic acid or a hydroxy aliphatic acid having 8 to 30 carbon atoms (straight or branched, saturated or unsaturated), (b) a straight or branched dibasic acid having 12 to 36 carbon atoms, and (c) glycerol or a glycerol condensate, and a hydrocarbon having a branched structure.

The invention relates to an end-to-end tumor segmentation method based on a multi-attention mechanism. The method mainly comprises a backbone network part and an attention module part, the backbone network comprises three sub-networks, and the three sub-networks are composed of improved 3D Resinual U-net composition; and the attention module is composed of a specially designed double-branch structure. The method can make up the defect of low training efficiency in the prior art; defect of poor segmentation precision, converting the multi-class segmentation problem of a plurality of tumor sub-regions into a plurality of two-class segmentation tasks. The attention mechanism takes the segmentation result of the tumor peripheral edema region as a soft attention and adds the soft attention intothe segmentation subtask for the tumor core part, and the segmentation result of the tumor core part is also added into the segmentation subtask for the enhancement region in the tumor core through the attention mechanism. The method is suitable for segmentation of 3D images of tumor lesion tissues with similar hierarchical structures including brain tumors, including MRI images, CT images and the like, and a more accurate segmentation result can be provided.

This invention provides a dense, high-purity colloidal silica containing silica secondary particles having a branched and/or bent structure, and a production method thereof. Specifically, this invention provides a method for producing a colloidal silica, comprising the steps of 1) preparing a mother liquid containing an alkali catalyst and water, and having a pH of 9 to 12; and 2) adding a hydrolysis liquid obtained by hydrolysis of an alkyl silicate to the mother liquid, wherein the step of adding the hydrolysis liquid to the mother liquid sequentially comprises A) step 1 of adding the hydrolysis liquid until the pH of the resulting liquid mixture becomes less than 7; B) step 2 of adding an aqueous alkali solution until the pH of the liquid mixture becomes 7 or more; and C) step 3 of adding the hydrolysis liquid while maintaining the pH of the liquid mixture at 7 or more, and a colloidal silica containing silica secondary particles having a branched and/or bent structure, obtained by this method.