125results about How to "Fully generated" patented technology

The invention discloses a generative multi-round chat dialogue method and system and a computer-readable storage medium. The method is divided into two stages: a stage 1, training a multi-round chat dialogue model by using dialogues of a corpus; and a stage 2, inputting a question to be answered, posed by a user, into the trained multi-round chat dialogue model, and outputting an actual answer. Bymining keywords in historical dialogues and introducing an attention mechanism into the model, all words in the historical dialogues are treated differently, which expands the role of the keywords inthe historical dialogues in generating answers. A width channel is used to predict keywords and broaden topics, a depth channel is used to predict the weights of the keywords in the historical dialogues so as to penetrate into the current topic, and the topic information obtained from the two parts is introduced into a decoder to assist in decoding. The guidance of the topic information effectively solves the problem of meaningless answers, and greatly reduces the number of meaningless answer.

The invention relates to an adaptive generation system for image semantic description, and belongs to the technical field of image semantic description. The system comprises an image reader, an encoder, a decoder and a semantic description display. The image reader is used for acquiring an image to be described semantically. The encoder comprises a global feature extraction unit, a local feature extraction unit and an image feature combination unit. The decoder comprises a single-layer or multi-layer neural network. The neural network comprises an LSTM part, an Attention part and an MLP part,and a semantic description model is generated. The Attention part adopts a self-adaptive attention mechanism. The decoder generates words and sentences of image semantic description by utilizing the semantic description model according to the combination information output by the encoder. The semantic description display is used for outputting and displaying words and phrases of the image semanticdescription. The focus key point of the image is determined, higher-level semantic information is mined, and the detail information problem describing words or sentences is perfected.

The invention discloses a mode collapse resistant robust image generation method based on a novel conditional generative adversarial network. Compared with a method in the related art, the technologydisclosed by the invention has the advantages that the adaptability is strong, the robustness is good, only the category of a required image is required to be specified in the using phase, not manualintervention is required in the process, the training phase is short in consumed time, the training process is stable, and the balance between the diversity and the authenticity of generated images issufficiently maintained. The main innovation of the mode collapse resistant robust image generation method lies in that a mode collapse problem and a training failure problem in the training processof other conditional generative methods are solved. Meanwhile, parameters are respectively optimized for a classifier and a discriminator, thereby avoiding problems of instable training and mode collapse of the methods of the same category. In addition, the invention further introduces a construction strategy for weight sharing, so that the training speed is greatly improved and the storage overhead is reduced under the premise of not damaging the original performance. The mode collapse resistant robust image generation method is applied to a diversified image data generation task of low-costlarge-scale specified labels.

The invention relates to an image semantic description generation method, and belongs to the technical field of image semantic description. The method comprises the following steps: S1, extracting global features and local features of an image; S2, combining the global features and the local features to obtain combined information; and S3, focusing image features and focusing texts according to the combination information through an adaptive attention mechanism to generate a semantic description model, and generating words and phrases of image semantic description by utilizing the semantic description model. According to the method, the problems that the focus key point of the image is determined, higher-level semantic information is mined, and detail information describing words or sentences is perfected in the existing image semantic description technology are solved.

The invention discloses a preparation method of a high-yield high-activity alkynol-coated vinyl platinum complex. The preparation method is characterized by comprising the following steps that 1, 100 parts by weight of 1,3-divinyltetramethyldisiloxane is taken to be poured into a reaction kettle; 2, 1-8 parts by weight of chloroplatinic acid is dissolved with solvent to obtain a chloroplatinic acid solution, the chloroplatinic acid solution is poured into the reaction kettle to be stirred, the temperature is increased to 50-60 DEG C, and the reaction time is set for 0.5-2 h; 3, 5-30 parts by weight of sodium bicarbonate is added into the reaction kettle, then the temperature is increased to 60-80 DEG C, and stirring is continuously performed to enable a reaction to be performed for 1-3; 4, 1-10 parts by weight of alkynol is dissolved with 10-100 parts by weight of solvent to obtain an alkynol solution, the alkynol solution is dropwise added into the reaction kettle, and dropwise adding is completed in 0.5-2 h; 5, filtering is performed to obtain filtrate and sodium bicarbonate residues; 6, reduced-pressure solvent removing is performed on the filtrate at 60 DEG C to obtain liquid; 7, vinyl silicone oil is added into the liquid to obtain the high-yield high-activity alkynol-coated vinyl platinum complex. The preparation method is simple and easy to achieve, and the yield of the prepared product can be more than 98 percent.

The invention relates to an optical remote sensing image saliency target detection method, and the method comprises the steps: S1, building a dense attention flow network which comprises an attentionflow guided feature coding module and a progressive feature decoding module; s2, importing an optical remote sensing image; s3, processing the optical remote sensing image by adopting an attention flow guided feature coding module to generate enhanced features with higher discrimination capability, the attention flow guided feature coding module mainly comprising a global context perception attention module and a dense attention flow structure, and S4, decoding the enhanced features in the step S3 by adopting a progressive feature decoding module, gradually fusing the deep features and the shallow features in a feature decoding stage, and generating a plurality of side outputs and final outputs under the supervision of the saliency map and the saliency edge map.

The present application discloses a technology of radio frequency plasma treatment using radio frequency waves of two different frequencies, in which plasma can be sufficiently and stably generated and maintained. The first frequency is used to generate a plasma by inducing a discharge, and the second frequency is used to generate a self-bias voltage on the substrate being processed. The wave of the second frequency is applied with a time lag after the application of the radio frequency wave of the first frequency. The present application also discloses impedance matching techniques in radio frequency plasma processing in which the impedance provided is optimized for inducing discharge and stabilizing the plasma.

The invention discloses a graphical user interface generating method and a device of the same, wherein, according to GUI generated parameters which are preset and comprise congruent relationship of a register and a graphical user interface GUI control, and a preset GUI control module is applied to generate a GUI control corresponding to a register of a register list; the GUI generating method and the device display the generated GUI control on the GUI interface, which realizes the automatic generation of GUI, improves the generation efficiency of GUI and avoids the probable mistakes caused by manual programming; besides, the preset generation parameters can realize a plurality of GUI generation methods, which makes the generation of GUI more flexible and diverse.

The invention provides an image generation method and system based on a capsule network. The method comprises the following steps: firstly, constructing training data according to attribute characteristics of an image to be generated; Constructing an image generation model; Training the constructed image generation model by adopting a batch training method according to the training data to obtaina trained image generation model, and constructing a random noise vector; And finally, taking the attribute vector and the random noise vector of the to-be-generated image as the input of the trainedimage generation model, and generating a new image with the same size as the image in the training data. The method has the beneficial effects that the capsule network is added into the image generation model to serve as the encoder network, so that the faster convergence of the model training process is facilitated; And on the other hand, compared with the pooling process of the convolutional neural network, the dynamic routing algorithm of the capsule network has stronger generalization robustness on characteristics, and more diversified and real images can be generated.

The invention discloses a method for preparing magnesium nitride powder. The method comprises the following steps of: adding nano magnesium powder into a quartz tube; and fully reacting the nano magnesium powder with high-purity nitrogen at the temperature of between 600 and 900 DEG C to prepare the nanoscale magnesium nitride powder. The preparation method is simple, the problem that magnesium is easily oxidized by air in the conventional preparation method is solved, and high-purity magnesium nitride powder is prepared.

The invention relates to a method for preparing an aluminum alloy component through a powder metallurgy method, and belongs to the technical field of preparing aluminum alloy components through the powder metallurgy method. According to the preparation method, after pure Al powder and three kinds of alloy element powder or binary alloy powder of Mg, Si and Cu are mixed uniformly according to a certain proportion, the aluminum alloy component is obtained through the processes of press forming, sintering, heat treatment, subsequent surface treatment and the like. According to the method, the pure aluminum powder is use as the main raw material, and an aluminum alloy system suitable for a pressing-sintering process is developed; and the sintering-heat treatment integrated production process is studied, the production process is simplified, the production efficiency is improved, and the production cost is reduced. The density of an aluminum alloy material prepared through the method is higher than 98%, the tensile strength of the aluminum alloy material is larger than 300 MPa, and the ductility of the aluminum alloy material is about 5%.

The invention provides uracil heating synthesis method. The method comprises the following step that materials including oleum containing sulfur trioxide with the mass percent being 8 to 12 percent, urea and malic acid are proportioned according to the mass ratio of (30 to 40):(1 to 5):(3 to 7); under the stirring condition, the urea is firstly added into the oleum; then, the malic acid is added into the oleum to obtain a mixed material; the mixed material is subjected to microwave heating, so that the mixed material is subjected to condensation reaction at the temperature of 85 to 90 DEG C; acondensation material is obtained; the condensation material is mixed with water; filtering is performed to obtain a uracil crude product. The heating synthesis method provided by the invention has the advantages that the steps are simple; the consumed time is short; the yield of the produced uracil product is high; the production cost is low; the pollution is reduced.

The invention discloses a Ti-B-C-N ceramic coating and a preparation method thereof. The Ti-B-C-N ceramic coating is prepared from the following components: 35% to 45% of Ti powder, 30% to 40% of B4C powder, 8% to 15% of graphite powder, 10% to 15% of Co powder and 1% to 8% of adhesion agent PVA (Poly Vinyl Acetate). The method comprises the following steps: firstly, preparing composite powder required for the spraying of the ceramic coating; then preparing the Ti-B-C-N ceramic coating by using a reaction plasma spraying method. According to the method, the prepared Ti-B-C-N ceramic coating comprises TiN, C0.3N0.7Ti and TiB2 which are taken as main phases, and also contains a small quantity of BN and other materials. According to the prepared Ti-B-C-N ceramic coating, the problem of poor coating performance due to the less content of B-C in the Ti-B-C-N coating prepared by using the conventional reaction plasma spraying method is solved. In addition, oxides are nearly not contained in the Ti-B-C-N ceramic coating; the hardness of the coating is improved due to BN existing in the coating.

The invention discloses a hydrogenation catalyst composition and a preparation method thereof. The hydrogenation catalyst composition comprises an aluminum oxide carrier, an active metal and an auxiliary agent, wherein the active metal is selected from one or more of VIII family and/or VIB family metallic elements, and the auxiliary agent is selected from one or two of silicon or zirconium; the concentration of silicon and/or zirconium in the catalyst, in the form of oxides, are gradually increased from centers to outer surfaces of catalyst particles, wherein the concentration of silicon and/or zirconium at a position of 1/3R of particles is 0.1-0.5wt%, the concentration of silicon and/or zirconium at a position of 2/3R is 0.5-1.0wt%, the concentration of silicon and/or zirconium at a position of R is 1.0-3.0wt%, and R is the semi-diameter of the catalyst particles by taking the center of the catalyst as an initial point. The catalyst prepared by adopting an impinging stream co-precipitation gel forming mode in the invention is centralized in particle size distribution, and has an ideal pore structure, and the auxiliary agent and the active metal are distributed in the catalyst in a gradient increase mode, so that the deactivation rate of the catalyst can be effectively reduced, and the desulfurization and denitrification capacities of the catalyst are improved.

The invention discloses a combined test device for high-temperature and high-pressure foam in-situ generation and evaluation and a use method. The device comprises a foam in-situ generation system, ameasurement system and a back pressure system. The foam in-situ generation system comprises a gas transmission channel and a liquid transmission channel; the measuring system comprises a sand filled pipe, a high-pressure visual container and a capillary pipe; the back pressure system comprises a water storage intermediate container, a back pressure valve, a pump and a first water storage apparatus. After foams are generated in situ, the real state of the foams in the stratum can be observed more visually through the high-pressure visual container under the condition that the environment is notchanged, foam related data can be obtained accurately without any interference, and a more stable pressure environment is provided by additionally arranging the water storage intermediate container in front of the back pressure valve.

The invention provides an image generation method and a device based on a 3D rendering engine, The client sends the 2D user-defined graphics and graphics information to the 3D rendering engine, so that the processing logic for generating the user-defined images is placed outside the client, and the operation limitation of the client is broken through, and more diversified geometry generation is realized, which is convenient for the user to model based on the user-defined graphics.

The invention provides a method for enhancing point-edge interaction of a graph neural network.; obtaining a directed molecular graph G and graph structure data thereof; according to the graph structure data, obtaining all node original features Xv in all and the graph structure data according to all and all created node original features Xv in all and the graph structure data according to all andall created node original features Xv in all and the graph structure data; the graph neural network is iterated to the Kth layer, a final node representation form h (v) of the directed molecular graph is obtained, k is larger than or equal to 1, and K is larger than k; the hidden representation from the adjacent node w of each arbitrary node v to the edge of the arbitrary node v is utilized, namely, the message vector of the arbitrary node v on the kth layer is created, so that the information of the edge is associated and transmitted with the node information, the embedding of the nodes andthe edge is updated in the neural network training process, and the transmissibility of the information between the nodes and the edge is concerned.

The invention relates to a method for preparing a high-purity mullite monocrystal, in particular to the method for preparing the high-purity mullite monocrystal by utilizing a kyanite concentrate as a raw material at low temperature. The invention aims to provide the method for producing the mullite monocrystal with the advantages of energy conservation and consumption reduction, high yield, high purity, low temperature and environmental friendliness. The method is characterized by comprising the following steps of: (1) mixing kyanite concentrates of the same particle size with industrial Al2O3 in a ball grinder, wherein mass fractions of Al2O3 and SiO2 in a mixed sample accord with a mullite chemical structure; and (2) putting catalyst AlF3 outside the mixed sample, and carrying out heat treatment in two steps in a sealed high temperature resistant container: step 1, finishing nucleation of the mullite from the room temperature to a temperature of 900 DEG C; and step 2, at the temperature of 900-1100 DEG C, finishing crystallization of the mullite; and taking out after naturally furnace cooling to the room temperature to obtain the high-purity mullite monocrystal. Compared with the existing method, the method provided by the invention has the advantages of low synthesis temperature, high yield and purity, energy conservation and environmental friendliness, and is suitable for industrial production.

The invention discloses an urban path image generation method combining questionnaire survey and street view pictures, and the method comprises the steps: recording the start and end points of travelof residents and the feeling along the way in a given region range; capturing longitude and latitude coordinates of a starting point and an ending point and generating all travel paths; obtaining a main travel path of residents in a given territorial range by utilizing a linear density analysis method, and uniformly obtaining view finding points of street view pictures on the main path and tangentangles of the view finding points on the path by adopting an equidistant sampling method; dividing factors such as sky, buildings, afforestation and roads on the streetscape picture, calculating to obtain an average sky view rate, a building view rate, an afforestation view rate and a road view rate of the main path, and forming a space composition of the main path; integrating the main path space composition of the research area and the main path feeling score of residents to form an urban path image. Objective reality of a city and subjective feelings of people are comprehensively considered, and a research method of a city image is expanded.

The invention relates to welding methods for the fourth generation nuclear reactor key structural materials, in particular to a welding method applied to aluminiferous oxide dispersion strengthened ferrite / martensitic steel. According to the welding method, iron-based intermediate layer alloy foil sheets with components of (8-18)wt.% Ni, (8-18)wt.% Cr, 3wt.% B, 2wt.% Si and (0.05-0.08) wt.% C are adopted, a transient liquid phase diffusion bonding method is used for welding the aluminiferous oxide dispersion strengthened ferrite / martensitic steel, the welding temperature is 1150-1200 DEG C, the thermal insulation time is 1-2h, the pressure is 5-10MPa, an NiAl phase with high density and a nanometer size is generated at the interface of welded aluminiferous dispersion strengthened ferrite / martensitic steel and intermediate layer alloy foil, and tensile samples of the welded aluminiferous dispersion strengthened ferrite / martensitic steel with a range from the room temperature to700 EDG C all rupture at the welded aluminiferous dispersion strengthened ferrite / martensitic steel part.