148results about How to "Generate good effect" patented technology

The invention relates to high-efficiency hydrate inhibitive environment-friendly drilling fluid, and particularly relates to the drilling fluid containing water, bentonite, hyamine dimeric surfactant, inorganic salt, sodium carboxymethyl starch, cationic polyacrylamide and solid anti-sloughing lubricating agent GFRH. The high-efficiency hydrate inhibitive environment-friendly drilling fluid can be used for exploiting continental or ocean oil-gas resources and can effectively inhibit the generation of hydrate; moreover, the components of the drilling fluid are small in environment harm, and the drilling fluid has the advantages of high efficiency and environmental protection.

The invention discloses establishment of a virtual case library of cancer pathological images and a multi-scale cancer detection system based on a convolutional neural network. The system is based ona method of a convolution neural network, a cancer mass region is detected on a pathological full-scan section, and the system includes four modules: 1) a pathological section image preprocessing module; 2) a virtual case database construction module; 3) a high-scale cancer mass detection module; and 4) a small-scale cancer mass classification module. The multi-scale cancer detection system provided by the invention can make full use of the multi-scale information of a pathological image, on different scales, according to the characteristics of the image, different strategies are designed to detect a suspected cancer area, and at the same time, under the condition of insufficient training data, the virtual case library method established by the invention can provide more training data setsfor an existing data-driven deep learning method. The multi-scale cancer detection system based on a convolutional neural network has the characteristics of multi-scale detection, driving of a relatively small amount of data and the like, and has the characteristic of reducing computing resources required for one-time recognition and improving time efficiency of an algorithm on the basis of ensuring the overall recall rate and accuracy.

The present invention discloses a method of reconstructing a three-dimensional dental model based on photos or videos, and a method of reconstructing the three-dimensional dental model based on RGBD images, and relates to the field of image processing. The first method is composed of an incremental approach and a global approach. According to the incremental approach, in the condition that when a generated dental model is not complete enough, or a certain key area of the generated dental model is not clear enough, photos shot at new angles are additionally provided and subjected to treatment, so that a better generating effect is realized. According to the global approach, in the condition that the key areas of the dental model can be clearly expressed when shot at fixed and complete shooting angles, the dental model is generated through one shot without the manual determination or the manual intervention. According to the RGBD approach, the depth information does not need to be calculated, so that the calculation of non-RGBD images is optimized. When the depth information is calculated, the accuracy of the acquired depth information based on RGBD images is higher than that of to-be-estimated non-RGBD images. According to the technical scheme of the invention, corresponding optimization schemes are respectively provided for the above three approaches. Therefore, process flows are reduced and the processing accuracy is improved.

The invention discloses an exterior wall coating material capable of finishing working through one-time painting.The exterior wall coating material is prepared from natural resin, polyvinyl acetate emulsion, deionized water, quartz powder, zinc oxide, sulfonated hyperbranched polypyrrolone modified graphene oxide, polybenzimidazole modified meso-porous silicon, KH-560 and inorganic pigment.The exterior wall coating material capable of finishing working through one-time painting has the better adhesive capacity, aging resistance, covering power, washing resistance, scratching resistance, stain resistance, self-cleaning performance, fastness, acid resistance and alkali resistance, longer service life and the like compared with existing water-based exterior wall coating materials.

The present invention relates to a half-body bath / sauna machine for activating blood circulation and the metabolism by warming the lower half of the human body by means of steam, and more particularly, to a steam warming half-body bath / sauna machine for warming the lower half of the human body with medicated steam to prevent and treat blood circulation disorders, metabolic functional disorders, gynecological disorders, stress, poor circulation, geriatric diseases, skin disorders and the like, by providing a sauna compartment that enables a user to sit comfortably with the lower half of the body (the legs) straightened, while medicated steam is uniformly fed into the sauna compartment from a steam device through a steam tube, and providing a cover panel integrally formed with an ingress / egress door of the sauna compartment. In addition, the steam warming half-body bath / sauna machine of the present invention is provided with said sauna compartment that enables a user to sit comfortably with the lower half of the body (the legs) straightened, a steam device configured for good steam generation, and a steam tube, which enable the uniform feeding of medicated steam into the sauna compartment, and the amount of steam generated by the steam device can be suitably controlled through precisely controlling the amount of water supplied to the steam device.

A chemical for removing pollutants from drinking water contains perferrite, permanganate and assistant. It features strong catalytic oxidizing action for destroying oraganic substances.

The invention relates to a preparation method of a catalyst for the selective hydrogenation of alkyne and alkadiene. A carrier of the catalyst is a compound of 30-99% of aluminum oxide, 1-50% of titanium oxide, 1-30% of magnetism oxide and 1-40% of zinc oxide based on the weight of the carrier of the catalyst as 100%; meanwhile, an alcohol reagent and an alkali metal or an alkaline-earth metal are used as modifying reagents; the preparation method is an acidity sol mixed kneading molding method; the active components of the catalyst comprise three or more of palladium, gold, silver, bismuth, cerium, lanthanum, tin and antimony, and a loaded solution of the active components is subjected to polymerization; the carrier of the catalyst prepared by the method has relatively low acid content; and the prepared catalyst shows up relatively appropriate reaction activity in the selective hydrogenation reaction of the alkyne and the alkadiene and has relatively good hydrogenation selectivity, relatively strong green oil generation resistance and relatively low palladium wastage rate.

The invention relates to a single-viewpoint video depth obtaining method based on scene classification and geometric dimension. The single-viewpoint video depth obtaining method specifically comprises the following steps: (1) judging whether a current frame image belongs to a scene in which a camera is static and an object moves or a camera moving scene; (2) judging whether needing to estimate an initial depth graph of the current frame image; (3) figuring out the initial depth graph of the current frame image; and (4) for the scene in which the camera is static and the object moves, obtaining a motion depth graph of the current frame image and fusing the motion depth graph of the current frame image with the initial depth graph; for the camera moving scene, carrying out global motion compensation, carrying out motion estimation on adjacent frames of images after the global motion compensation by using an optical flow method, judging whether a moving object exists, and determining whether needing to fuse with the initial depth graph. By adopting the single-viewpoint video depth obtaining method provided by the invention, no specific scene is relied on, the calculated amount is appropriate, the generated noise is small, a depth graph better conforming to actual scene distribution is obtained, and a 3D video with a better effect is synthesized.

The invention discloses a Chinese character font style migration method based on deep learning. The Chinese character font style migration method specifically comprises the following steps: 1, acquiring a Chinese character image of a specified input character set; 2, initializing the color, the background and the size of the font image; 3, performing learning training by taking the source font image x and the real images y of the five target fonts as priori knowledge through a generator G; 4, taking the source font image x as condition input by the generator G, using a down-sampling area to extract a shallow font feature F0, restoring the feature vector into an image through an up-sampling area, and generating a generated image G (x) of the target font; 5, discriminating and classifying the generated image G (x) of the target font by using a discriminator D; and 6, calculating the value of the loss function. The method overcomes the problems that an existing research method cannot fully extract detail features of fonts, and the generated fonts have structural deformation, detail loss and fuzzy contours.

The invention discloses a refined embroidery texture migration method based on a sample domain, which comprises the following steps: in a model training stage, based on a cyclic generative adversarialnetwork, adding a Waserstein loss item, a feature matching loss item and an MS-in loss item into a loss function, and performing feature matching on the loss item and the MS-in loss item; according to the SSIM loss item, a self-attention mechanism is added to a generator G network, and a resolution enhancement method from coarse to fine is introduced. A training model is used for synthesizing embroidery textures; performing color quantization processing based on self-organizing network mapping on the input embroidery vector design drawing, performing color correction on the generated embroidery textures through the quantization method, and finally combining the embroidery textures with the generated normal map to obtain an embroidery texture map which meets the three-dimensional display requirement and has both embroidery colors and concave-convex textures. According to the method, rapid generation from image design to refined embroidery texture is realized, and the rework rate of production personnel is reduced while the communication efficiency of the design personnel is improved.

The composite water pollutant eliminating chemical preparation consists of permanganate, ferrate, silicate polymer and adjuvant. The present invention has the efficient pollutant eliminating and flocculating effects of composite potassium permanganate-ferrate preparation as well as the adsorption, rolling and sweeping effects of inorganic polymer. The composite water pollutant eliminating chemical preparation contains natural sterilizer and the powerful oxidation property may be utilized to destruct viruses and bacteria, generating less side sterilizing product. In addition, the composite water pollutant eliminating chemical preparation has excellent raised coagulation and filtering effects for eliminating turbid matter, color, algae, bad smell, organic pollutant, mutagenic matter, etc.

The water pollutant eliminating chemical preparation consists of ferrate, silicate polymer and adjuvant. The water pollutant eliminating chemical preparation contains natural sterilizer and the powerful oxidation property may be utilized to destruct viruses and bacteria, generating less side sterilizing product. The present invention has raised coagulation and filtering effects for eliminating turbid matter, color, algae, bad smell, organic pollutant, mutagenic matter, etc.

The invention discloses a face image processing model training method and device, electronic equipment and a storage medium, and the method comprises the steps: carrying out the coding of a first faceimage through an encoder of a face image processing model, and obtaining a first feature vector of the first face image; splitting the first feature vector into a plurality of sub-feature vectors, wherein the sub-feature vectors at least comprise an identity sub-feature vector and an age sub-feature vector; determining a to-be-decoded vector according to each sub-feature vector, and decoding theto-be-decoded vector by using a decoder of the face image processing model to obtain a second face image; and optimizing parameters of the face image processing model according to the regression lossvalues of the first face image and the second face image. According to the face image processing model obtained through training, the dependence on data distribution is reduced, the method is more robust to long-tail data with unbalanced ages, and a face aging image with a better effect can be generated.

Disclosed is a method for preparing a lithium battery electrode material FeSe2 thin film through magnetron sputtering. The method comprises the steps that a, preparation of sputtering is conducted, a silicon chip selected as a substrate and a FeSe target selected as a sputtering target are installed in a magnetron sputtering chamber, and the distance between the sputtering target and the substrate is adjusted to be 5-7 cm; b, sputtering deposition is conducted, a vacuum chamber is vacuumized until the air pressure is below 2*10-4 Pa, then argon gas is pumped in, the sputtering air pressure, the sputtering power and the temperature of the substrate are adjusted, and a thin film is deposited on the silicon chip after the sputtering is conducted for a certain time; and c, post-annealing treatment is conducted, the silicon chip where the thin film is deposited and a selenium granular ball are sealed in a vacuum quartz tube with the air pressure being smaller than 1*10-2 Pa, then the quartz tube is positioned in a tubular furnace, the post-annealing treatment is conducted under protection of the argon gas, and the FeSe2 thin film is obtained. The method is capable of preparing the FeSe2 thin film with a large area and good properties, simple in preparation process, low in energy consumption, low in cost, high in efficiency, good in repeatability, and suitable for industrialized production.

The invention discloses a brand-new molecule generation method based on a multi-task capsule auto-encoder neural network. The method comprises steps of building a brand-new molecule generation model which comprises an encoder, a multi-task capsule classifier and a decoder through an auto-encoder frame; expressing drug molecules as SMILES (simplified molecular linear input specification), marking atarget property label, and learning characteristics of known drug molecules through a training stage to obtain a training model; reconstructing molecules by utilizing the training model through a reconstruction stage; in the generation stage, molecules being generated by using a training model, the generated molecules having multiple set target properties at the same time, and meanwhile, the generated molecules having a large number of new molecules and new skeletons. The method can be used for generating various molecules such as medicines or compounds, and the characteristics and propertiesof the known medicines can be learned through one-time training, so the molecules meeting the required physical, chemical and biological properties at the same time are generated; the molecules generated by the method are higher in effectiveness and more excellent in property.

The invention discloses a code automatic generation device and method based on image target detection, and the device comprises a control element target detection module which is used for carrying outthe feature extraction of an input front-end image, generating a candidate region, carrying out the target classification, and carrying out the frame regression fine trimming of a position, so as toobtain a final accurate position; The rule-based DSL code generation module is used for sequencing the control elements in the horizontal direction and the vertical direction according to the positioninformation of the control elements output by the control element target detection module, and generating DSL codes by formulating rules; And the compiler module is used for compiling the DSL codes generated by the rule-based DSL code generation module into real front-end codes, so that a more accurate generation effect can be achieved, and a new research idea is provided for the descriptive programming automation field.

The invention provides a recommendation model training method, and the method comprises the steps: carrying out the training of a training sample set through employing a single-class domain perceptionlow-rank model, so as to obtain a model parameter matrix, and enabling the model parameter matrix to be used for generating a recommendation model. The scheme of the invention can be applied to the field of artificial intelligence recommendation. Various domain characteristics and positive and negative samples can be introduced into the single-class domain perception low-rank model, so that morefactors related to user selection and negative example information related to the user can be considered in the training process, a recommendation model with more excellent performance is generated, and the recommendation result better conforms to user requirements. In addition, the specific training mode of model training can greatly simplify the complexity of model training, so that the model training efficiency is improved, and introduction of more information becomes possible.

The invention relates to the field of drilling fluids for petroleum, natural gas and natural gas hydrate drilling and discloses a water-based clay-free drilling fluid and application of the drilling fluid to polar region frozen soil stratigraphical drilling. The drilling fluid comprises an ultralow-temperature base liquid, a tackifier, a filtrate reducer, a blocking agent, a lubricating agent andwater, wherein the ultralow-temperature base liquid is composed of inorganic salt and alcohol, the solidifying point of the ultralow-temperature base liquid is lower than or equal to -28 DEG C, and the heat conduction coefficient is smaller than 0.5W / m.K. The water-based clay-free drilling fluid has good rheological properties, filtration properties, low heat conduction characteristics, well wallstability and environment-friendliness even under the condition of -25 DEG C and can be applied to polar region frozen soil stratigraphical drilling.

The invention relates to a method for generating distractors of English similar word forms by being combined with word class. The method includes steps of selecting a source word from a word bank as a source word character string, utilizing other words as target word character strings, traversing all words in the word bank, and solving similarity between the source word character string and the target character strings according to uniformized integration similarity algorithm; controlling the threshold value of the similarity within 0.6-1.0, and taking the words within the range of the threshold value as optional words; subjecting the optional words and the source word output in the last step to similarity calculation combined with the word class, and controlling the threshold value a of the similarity within 0.6-1.0, thereby obtaining the distractors of the source word; finishing once processing course. By introducing the LCS (longest common subsequence) algorithm to uniformized integration, blindness in calculating similarity of the English words by singly depending on one similarity algorithm is changed, reliability and accuracy in generation of the distractors of the English similar word forms are improved, and the problem that words with same meaning but in different word classes repeatedly appear is solved.

The disclosure relates to a photograph synthetic method and apparatus. The method comprises the following steps: after an instruction to generate a synthesized photograph is received, starting an image acquisition part to acquire a photograph; when a current photograph is acquired, calculating the expression score of a first type of faces in the current photograph, wherein the first type of faces are indicative of the faces the expression scores of which calculated through the photographs acquired before the current photograph are not greater than a preset score threshold; determining whether the expression score of each of the first type of faces in the current photograph is greater than the preset score threshold; if all the expression scores are greater than the preset score threshold, controlling the image acquisition part to stop acquiring the photograph, and generating the synthesized photograph by splicing a second type of faces in the acquired photograph. The method and apparatus can reduce the number of generated synthesized photographs on the basis of a good synthetic photograph effect, thereby effectively shortening the time for generating the synthesized photograph and reducing the power consumption of generating the synthesized photograph.

The invention relates to a two-stage expression animation generation method based on double generative adversarial networks. The method comprises the following steps of: firstly, extracting expressionfeatures in a target expression profile diagram by utilizing an expression migration network FaceGAN in a first stage, and migrating the expression features to a source face to generate a first-stageprediction graph; in a second stage, supplementing and enriching the details of eye and mouth regions which contribute much to expression changes in the prediction graph in the first stage by using adetail generation network FineGAN; and generating a fine-grained second-stage prediction graph and synthesizing a face video animation, wherein an expression migration network FaceGAN and the detailgeneration network FineGAN are both realized by adopting a generative adversarial network. According to the method, the two-stage generative adversarial network is provided for expression animation generation, expression conversion is carried out in the first stage, image detail optimization is carried out in the second stage, the designated area of the image is extracted through a mask vector, emphasis optimization is carried out, and meanwhile, a local discriminator is combined for use, so the generation effect of important parts is better.

The invention discloses a video dense event description method based on a generative adversarial network, and belongs to the field of deep learning and image recognition. The method comprises the steps of constructing a video dense event description network; the network comprises a video feature extraction module for performing feature extraction on video frames to obtain video frame features; thetime sequence action feature extraction module is used for performing video frame feature learning by utilizing the characteristics of forward propagation and reverse propagation of the video to obtain each time sequence action feature; the natural language encoder fuses the video frame features and the time sequence action features by utilizing an attention mechanism to obtain natural statements; the discriminator is used for enhancing the accuracy of natural statements; and performing video dense event description by using the trained video dense event description model. According to the method, the characteristic of video two-way propagation is fully considered, video characteristics and time sequence action characteristics are fully utilized when the natural sentences are learned to be generated, and the grammar discriminator and the content discriminator are constructed, so that the accuracy of the natural sentences is effectively enhanced.

The invention relates to a face age synthesis method and system, and relates to the technical field of face synthesis, and the method comprises the steps of carrying out the preprocessing of a test image, inputting the test image into a face age synthesis model, and obtaining face images of different age groups, wherein the training of the face age synthesis model comprises the step of optimizingparameters of a generator network by utilizing a loss function of the generator network according to an attention mask and a synthesized face image, taking the synthesized face image and the corresponding original input as negative samples; taking randomly selected face images in the randomly generated target age labels during synthesis as positive samples and inputting same into the discriminatornetwork to obtain the authenticity probabilities of the corresponding face images, and optimizing parameters of the discriminator network through a loss function of the discriminator network according to the probabilities; and when the number of iterations reaches the maximum number of iterations, obtaining a trained face age synthesis model. According to the invention, the visual quality of faceage synthesis can be improved.

The invention discloses a text generation method and device, equipment and a storage medium, and belongs to the field of data processing. The method includes: obtaining a first text; obtaining semantic features of the first text based on the characters in the first text. That is to say, in the process, semantic feature extraction is performed by taking the character of the first text as a unit, the semantic features reflect the relationship between the characters to a certain extent, and the semantic feature extraction is more accurate. Based on the randomly generated hidden features and the semantic features of the first text, probabilities corresponding to the plurality of alternative characters are obtained. In this way, texts with different expression modes can be generated, and the text generation diversity is improved. A plurality of second texts are generated based on the alternative characters of which the probabilities meet the target probability condition. Through the technical scheme, the server can improve the diversity of the generated text through the hidden features on the premise of ensuring the extraction accuracy of the semantic features of the first text, and a better text generation effect is achieved.

The invention discloses a whitening and anti-fatigue body lotion. The whitening and anti-fatigue body lotion of the invention contains vine tea extract, sweet tea extract and black medicine extract in the weight ratio of (0.2-2): (0.4-2): (0.2~2) mixed plant composition, the whitening anti-fatigue body lotion of the present invention achieves the effect of anti-fatigue by reducing the content of urea nitrogen, has the effects of anti-fatigue and anti-aging, and has good inhibition on tyrosinase Effect, through the inhibition of tyrosinase production to reduce melanin production to achieve whitening effect.

The invention discloses a PCB defect data generation method based on deep learning. In a generator network and a discriminator network, a BatchNorm layer, a Relu layer and a LeakyRelu layer are added,the Relu layer is used for the generator network, and the LeakyRelu layer is used for the discriminator network: an encoder is designed and is composed of three convolution layers, and the encoder extracts features from an input image; a converter is designed, wherein the converter is composed of six residual blocks and converts the feature vector from a source domain to a target domain; a decoder is designed, wherein the decoder is composed of three deconvolution layers; designing a discriminator, wherein the discriminator is composed of five convolution layers; a loss function is designed,wherein the loss function comprises four parts; a training set for model training is prepared, and a network is trained to automatically repair defects; a weight file under the optimal iteration frequency is selected for testing; the restored image and the test image are compared by using the LBP, so that the position of the defect can be found more accurately; and various defect data sets are obtained by utilizing the generated samples. The method is high in robustness, wide in application range and excellent in generation effect.

The invention discloses a map generation and navigation obstacle avoidance method based on an automatic driving platform. The method comprises the following steps of obtaining a space map of car driving based on an EKF improved Hector SLAM map algorithm; utilizing the obtained space map, and utilizing a navigation obstacle avoidance algorithm integrating A* and DWA to obtain a route plan of a trolley in the driving process in the space map; and evaluating and verifying the effects on a map generation algorithm and a navigation obstacle avoidance algorithm through experiments. The navigation obstacle avoidance method provided by the invention can be fused with existing different navigation obstacle avoidance algorithms, and has effectiveness, completeness and correctability.

The invention relates to the field of image processing, and provides a face image generation method and device, electronic equipment and a storage medium. The method comprises the steps of obtaining an original image including an initial expression of a target person and an expression image including a target expression; then, acquiring a mask image, a first image and a second image by a preset model according to the original image and the expression image, wherein the first image comprises five sense organ features of a target person matched with the target expression, and the second image comprises head features of the target person matched with the target expression; and fusing the first image and the second image according to the mask graph through a preset model to obtain a face image and a virtual expression of the target person in the face image, and matching the virtual expression with the target expression. Therefore, the face expression migration is realized, the facial features can be better matched based on the features of the five sense organs and the features of the head, the generated facial image can be more real, and the generated effect is better.

The invention discloses a cornu cervi pantotrichum external restoration composition. The cornu cervi pantotrichum external restoration composition is prepared from the following raw materials in parts by mass: 0.01 to 8 parts of cornu cervi pantotrichum, 0.05 to 8 parts of radix notoginseng, 0.5 to 8 parts of safflower; skin essence, a penetration enhancer and pharmaceutically accepted auxiliary materials also can be added. The cornu cervi pantotrichum external restoration composition has the advantages that symptoms are externally treated, and the congenital foundation is internally secured; the cornu cervi pantotrichum has the effect of enriching kidney essence to secure congenital foundation of a human body; the radix notoginseng and the safflower have the effects of tonifying Qi and blood, activating blood circulation and removing blood stasis, and promoting skin deep microcirculation, the essence element permeates into capillary blood vessels to participate in metabolism of the human body under the matching action of the skin penetration enhancer, the five internal organs are nourished to nourish interior by securing foundation and then nourish exterior by interior, and the functions of promoting restoration and nourishing skin are realized; the essence of the cornu cervi pantotrichum has the effects of restoring cells, healing wounds, improving microcirculation and the like; plant essences of the radix notoginseng and the safflower have the functions of fading spots, whitening and the like, and directly act on skin surface cells, so as to achieve the functions of nourishing, restoring, and treating symptoms by external applying; by compounding with first-class geo-authentic crude drugs such as herba centella, radix ginseng, Chinese angelica root and liquorice root, the effect of external applying is promoted, and the restoration capability is improved.

The invention relates to the technical field of livestock breeding, and specifically relates to a low-proportion roughage diet capable of promoting weight gaining of fattening lambs and beneficial tothe rumen health of the lambs. The roughage diet is prepared from the following raw materials in parts by weight based on dry substances: 10-25 parts of corn stalk, 2-6.2 parts of beet pulp, 32 partsof corn, 3.3 parts of molasses, 4-6 parts of soybean meal, 4-6 parts of cottonseed meal, 12-14 parts of corn germ meal, 4.5-25.6 parts of corn white rind, 0.8-0.9 part of stone powder, 0.8 part of gypsum powder, 1 part of bentonite, 0.6 part of table salt, 0.8 part of expanded urea and 1 part of premix. According to the roughage diet, the corn stalk is used as a main material, and a relatively lower level of roughage provides enough fibers for the growth of mutton sheep, so that the production performance of animals can be improved and the rumen health of the animals can be maintained.