264 results about "Layer group" patented technology

A fringe field switching thin film transistor substrate includes a gate electrode connected to the gate line, a source electrode connected to the data line, a drain electrode opposed to the pixel electrode and a semiconductor layer defining a channel between the source electrode and the drain electrode. A common electrode extends from the common line into the pixel area. A pixel electrode extends from the drain electrode into the pixel area overlapping the common electrode with the gate insulating film. The gate line and the common line are formed from a first conductive layer group having double conductive layers, and the common electrode is formed by an extension of the lowermost layer of the common line. The data line, the source electrode and the drain electrode are formed of a second conductive layer group having double conductive layers.

The invention discloses a multi-aspect deep learning expression-based image emotion classification method. The method comprises the following steps of: (1) designing an image emotion classification model: the image emotion classification model comprises a parallel convolutional neural network model and a support vector machine classifier which is used for carrying out decision fusion on network features; (2) designing a parallel convolutional neural network structure: the parallel convolutional neural network structure comprises 5 networks with same a structure, and each network comprises 5 convolutional layer groups, a full connection layer and a softmax layer; (3) carrying out significant main body extraction and HSV format conversion on an original image; (4) training the convolutional neural network model; (5) fusing image emotion features learnt and expressed by the plurality of convolutional neural networks, and training the SVM classifier to carry out decision fusion on the image emotion features; and (6) classifying user images by using the trained image emotion classification model so as to realize image emotion classification. According to the method disclosed by the invention, the obtained image emotion classification result accords with the human emotion standard, and the judgement process is free of artificial participation, so that machine-based full-automatic image emotion classification is realized.

A LCD device includes a gate line on a substrate and a data line crossing the gate line to define a pixel area; a thin film transistor source and drain electrodes; a common line parallel to the gate line; a common electrode extended from the common line and a pixel electrode extending from the drain electrode wherein the gate line and the common line have a first conductive layer group having at least double conductive layers, and the common electrode is formed by an extension of at least one transparent conductive layer of the common line; and the gate line, the source electrode and the drain electrode have a second conductive layer group having at least double conductive layers, and the pixel electrode is formed by an extension of at least one transparent conductive layer of the drain electrode.

The embodiment of the invention discloses a panel, a panel assembly and a terminal. The panel includes a non-display area; and a light-transmitting ink layer group is arranged at the lower surface of the non-display area. The light-transmitting ink layer group includes at least one light-transmitting ink layer; and a shading ink layer is arranged at the lower surface of the light-transmitting ink layer and is provided with a light transmitting hole for transmitting a light signal by a proximity sensor and a light receiving hole for receiving a light signal by the proximity sensor. The light-transmitting ink layer covers and hides the light transmitting hole and the light receiving hole. Therefore, the manufacturing process for an intelligent terminal is simplified.

The surface roughness of a copper sheet is adjusted to have an arithmetic mean roughness Ra of 0.5 μm or more and 4.0 μm or less in a direction parallel to a sliding direction upon connection, a mean projection-depression interval RSm of 0.01 mm or more and 0.3 mm or less in the direction, a skewness Rsk of less than 0, and a protrusion peak portion height Rpk of 1 μm or less. Further, as a surface coating layer, a Sn coating layer group X observed as a plurality of parallel lines is included, and a Cu—Sn alloy coating layer is present adjacent to each side of each of Sn coating layers constituting the Sn coating layer group X. The maximum height roughness Rz is 10 μm or less in a direction of part insertion. At the time of stamping the copper sheet, the copper sheet is surface-roughened by pressing, thereby forming depressions observed as a plurality of parallel lines in the surface of the copper sheet. The copper sheet is then plated with Cu and Sn, followed by reflowing to complete the production.

A device having an electrode-insulator layer-group III nitride layer structure, wherein an interface between the insulator layer and the group III nitride semiconductor layer lies along a non-polar plane of the group III nitride semiconductor layer is provided.

A laminated film with 200 or more laminated layers which include at least layers comprising a resin A and layers comprising a resin B where, taking the layers comprising the resin A having a layer thickness from 10 nm to 1000 nm as a layer group L, the layer group L has layers La, Lb and Lc and, in the direction from one surface to the other, these are arranged in the order of the layer La, the layer Lb, and the layer Lc, the thicknesses of the layers La and Lc are the same and at least 30 layers, each layer comprising the resin A, are included between the layer La and the layer Lc, and the thickness of the layer La is at least 1.4-fold or more and 0.7-hold or less the thickness of the layer Lb.

The invention discloses a method for generating an instruction sequence and a method and a device for executing neural network operation. The method for generating the instruction sequence of the neural network comprises the following steps: acquiring attribute data of each layer of the neural network; determining at least one hierarchical group and splitting parameters thereof according to the attribute data of each layer of the neural network, the hierarchical group comprising at least one layer in the neural network, the splitting parameters comprising at least one predetermined dimension and a predetermined number of parts in to which each predetermined dimension is split; and generating an instruction sequence of the neural network according to the attribute data of each layer of theneural network and the splitting parameters of the at least one layer group. According to the embodiment of the invention, all the layers in the neural network are grouped to realize simultaneous splitting calculation of all the layers in the neural network, and under the condition that the capacity of the memory is limited, the efficiency and the processing capability of executing neural networkoperation by hardware can be improved on the premise that the hardware cost is not increased.

The invention, which relates to the electronic material field, discloses a flexible heat-radiation film and a method for manufacturing the same. The heat-radiation film comprises a carrier film, a protection film layer, and at least one heat radiation layer group arranged between the carrier film and the protection film layer. The heat radiation layer group includes a heat-radiation adhesive film layer for horizontal heat transferring and a heat-radiation metal film layer at least being superposed on one surface of the heat-radiation adhesive film layer ad being used for longitudinal heat transferring, so that the heat radiation layer group having bidirectional heat transferring is formed. The heat-radiation adhesive film layer is made from raw materials including magnetic high-thermal-conductivity powder and resin. For the heat-radiation metal film layer, a material source is processed by a physical means to form atoms or molecules and then the atoms or molecules are deposited on the surface of the heat-radiation adhesive film layer. According to the heat-radiation film, the operation becomes simple; and the formed film layers are light, thin and compact. The flexible heat-radiation film can be applied to various electronic products and communication equipment.

The invention discloses a design method of composite material laminates with equal thickness. The design method comprises the following steps: establishing an optimal object through an optimization model; constructing a basic ply stacking structure meeting an engineering constraint according to an initial ply group sequence set in the performance optimization of the first stage; with the reference of the initial ply layer group sequence set in the first stage, sequentially adding plies under a principle of enabling D11, D22, D12 and D66 to be maximized according to the basic ply stacking structure, and adding plies of all the stacking corners into the basic ply stacking structure, so as to obtain the laminate ply stacking sequence basically meeting the engineering constraint; eliminating the plies stacked, and detecting whether the laminates subjected to ply elimination meet the performance constraint of the optimization model or not, so as to obtain the optimized ply stacking sequence of the laminates. With adoption the theoretic design method disclosed by the invention, designers can solve the complicated optimizing problem of the laminate ply stacking sequence, and can quickly find out the optimal solution of the laminate ply stacking sequence.

The invention discloses a preparation method of a group III nitride substrate. The preparation method comprises the following steps: selecting a silicon carbide substrate as a matrix, at certain vacuum degree, heating a SiC substrate to 1400-1650 DEG C through RF (radio frequency), introducing Ar/H2 mixed gas as a carrier gas in an epitaxy thermal decomposition process, and epitaxially growing multiple graphene layers on a carbon surface or a silicon surface of the SiC substrate; depositing and growing a group III nitride nucleating layer on the surface of multiple graphene layers; growing thick-layer group III nitride on the surface of the group III nitride nucleating layer; stripping the SiC matrix by adopting a mechanical stripping method, and grinding and polishing the surface of a group III nitride slice with the SiC matrix stripped, so as to obtain the group III nitride substrate. Multiple graphene layers which are large in area and high in quality can epitaxially grow on the carbon surface or silicon surface of the SiC substrate, and a high-quality group III nitride nucleating layer grows. The obtained nitride substrate slice is small in stress, low in defect density and high in crystal quality.

The present invention relates to a remote sensing image contrastive analysis method and system. The method comprises: S1: carrying out identification and segmentation on surface features in two remote sensing images that are taken at different times in the same area through the full convolution network, and obtaining segmentation images of all surface features in two remote sensing images, wherein the full convolution network comprises a plurality of convolution layer groups and a plurality of deconvolution layers, and the convolution layer group comprises convolution layers and parse convolution layers that are arranged alternately; and S2: carrying out contrastive analysis on the segmentation images of the same surface feature in the two remote sensing images so as to obtain a contrastive analysis result. The beneficial effects of the method and the system disclosed by the present invention are that: the technical scheme has good fault tolerance of the interference factors such as the atmosphere, the season, and the like, and also has a relatively high identification rate of dense surface features.

The invention discloses a generation method and device of a page layout file. The method comprises the following steps: obtaining a target image file which comprises a plurality of layers, and the grouping information of the plurality of layers, wherein the grouping information is used for indicating a layer group to which each layer belongs, and an organizational relationship between the layer groups; and according to a layout relationship between a layout area independently corresponding to each layer group and the layout area independently corresponding to each layer group, a control attribute independently corresponding to each layer and layout information, generating a page layout file, wherein the layout area independently corresponding to each layer group is determined by the layout information of the layers independently contained in each layer group, the layout relationship between the layout areas independently corresponding to each layer group is determined by the organizational relationship between the layer groups, and the control attribute independently corresponding to each layer is appointed by an input configuration operation. The scheme can improve GUI (Graphical User Interface) development efficiency and saves development cost.

The invention discloses is a mapping method of codeword flow-to-layer in MIMO system, which includes that the current layer of a MIMO system is divided into groups according to current number of codeword flow, and each codeword flow is mapped to the layer of a corresponding group. The invention also discloses a mapping device of codeword flow-to-layer in MIMO system, which comprises a layer grouping unit and a mapping unit; wherein, the layer grouping unit is used for dividing current layer of a MIMO system into groups according to the number of current codeword flow; the mapping unit is used for mapping each codeword flow to the layer of a corresponding group. The invention has simple implementation and more reasonable codeword flow-to-layer mapping mode.

The invention discloses a two-dimensional convolution operation processing method, system and equipment and a computer medium. The method comprises the following steps of reading to-be-calculated dataand convolution kernel operation parameters from a preset memory; calling a first matrix conversion processor to convert the to-be-calculated data into target to-be-calculated data in a matrix form,and meanwhile calling a second matrix conversion processor to convert the convolution kernel operation parameters into target convolution kernel operation parameters in the matrix form; and determining a target convolutional layer in a preset convolutional layer group, and transmitting the target to-be-calculated data and the target convolutional kernel operation parameters to the target convolutional layer, so that the target convolutional layer performs two-dimensional convolutional operation on the target to-be-calculated data and the target convolutional kernel operation parameters, and the number of the same convolutional layers in the convolutional layer group is greater than or equal to 2. According to the two-dimensional convolution operation processing method provided by the invention, the efficiency of two-dimensional convolution operation can be improved. The two-dimensional convolution operation processing system and device and the computer readable storage medium also solve the corresponding technical problems.

The invention discloses a method and a device for transmitting group call conversation information. The method comprises the steps that: during a beginning process of a group call conversation, a group call temporary identification is distributed for a called group by base station equipment; and the group call temporary identification and a corresponding non-access-layer NAS (Network Attached Storage) layer group identification of the called group are transmitted by the base station equipment. According to the method and the device for transmitting the group call conversation information which are disclosed by the embodiment of the invention, by adopting the advantage of a high bandwidth performance of an LTE (Long Term Evolution) system, the data transmission from a single neighborhood point to multi points is realized, the requirement on high use ratio of a hollow resource can be met under a trunking communication system, and thus the use ratio of the hollow resource can be increased.

The invention discloses a micro LED chip and a display panel. The micro LED chip comprises a substrate, a light emitting layer group positioned on the substrate, a first electrode and a second electrode, wherein the light emitting layer group at least comprises a first type semiconductor layer, a light emitting layer and a second type semiconductor layer which are arranged in a stacked mode; the first electrode and the second electrode are located on one side, far away from the substrate, of the light emitting layer group; the first electrode is electrically connected with the first type semiconductor layer; the second electrode is electrically connected with the second type semiconductor layer, wherein the first electrode is an annular electrode; the pattern of the outer edge and the pattern of the inner edge of the first electrode are the same; and the second electrode is located in the middle of a region surrounded by the inner edge of the first electrode. According to the embodiments of the invention, the electrode structure of the micro LED chip is improved, so that the difficulty of huge transfer of the micro LED chip is lowered.

The invention discloses a layer rendering processing method of a video image and a system thereof. The method comprises the following steps of reading multi-layer data of a current video image frame and carrying out grouping processing on the read multi-layer data to obtain one or more layer groups, wherein the layer data in the same layer group is not mutually exclusive, and whether the layer data is mutually exclusive is at least partially based on whether the layer data in the layer groups is overlapped or is at least partially determined based on a floating point value corresponding to a single channel color of the overlapped layer data before blending and a floating point value corresponding to the single channel color of the overlapped layer data after blending; carrying out pixel parallel rendering processing on the one or more layer groups. By using the method and the system of the invention, target maximization of visual elegance, design flexibility, and real-time processing can be realized.

Provided are a method of defining a focus movement order and a method of moving a focus. The method of moving the focus, according to the present invention, includes multi-dimensionally grouping two or more focusable elements into two or more focusable groups; displaying the two or more focusable groups; receiving a tab input from a user; moving the focus to another focusable group according to the tab input; receiving a layer change request from the user; moving the focus to a focusable element or a lower-layer group, which are included in a currently focused group; receiving a tab input from the user; and moving the focus to another focusable element or a lower-layer group according to the tab input.