215results about How to "Training accurately" patented technology

Embodiments of the present invention are related to a method and device for the determination and calculation of the depth of chest compressions during the administration of cardiopulmonary resuscitation (CPR). Embodiments use an optical sensor to monitor the distance that a victim's chest is displaced during each compression throughout the administration of CPR. The optical sensor is most commonly an image sensor such as a CMOS or CCD sensor, and more specifically a CMOS image sensor capable of three-dimensional imaging based on the time-of-flight principle. An infrared emitter may illuminate the victim's body and any visible piece of ground beside the victim. As the infrared light interacts with any surfaces it encounters, it is reflected and returns to the image sensor where the time of flight of the infrared light is calculated for every pixel in the image sensor. The distance data is used to gauge the effective displacement of the victim's chest. The optical sensors can be used to visualize the size of a patient and immediately gauge the body type and instruct the user accordingly. Furthermore, optical measurement techniques can be used to accurately measure chest rise during artificial respiration and ensure that proper ventilation is being administered in between compressions. In addition, optical measurements of the chest of the victim and the hands of the rescuer can be used to help ensure that the rescuer has positioned his or her hands in the anatomically correct location for effective CPR.

The invention discloses a large-data-flow-based network traffic abnormality detection system and method. The large-data-flow-based network traffic abnormality detection method is characterized by comprising the following steps: acquiring network packet information in network equipment in real time in a distributed manner; transmitting the network packet information to a distributed flow processing platform in real time for network data analysis, feature matching and access counting; storing the analyzed and detected network data into a large data platform according to an abnormal status so as to facilitate clustering analysis and classified training of the network data and dynamically update a network data protocol characteristic library. Through the system and the method, real-time detection is achieved through a distributed flow-type processing mechanism; through distributed storage of the large data platform and through the data calculating and analyzing capability, the distributed storage of the network data is achieved and the network data protocol characteristic library can be trained more accurately.

The invention discloses a sewage-disposal soft measurement method on the basis of an integrated neural network, and belongs to the field of sewage disposal. A sewage disposal process is high in nonlinearity, time-varying characteristics and complexity, and measurement for key water quality indexes is crucially significant in control of water pollution. In order to improve precision of simultaneous soft measurement for various key water quality parameters in a sewage-disposal soft measurement process by the sewage-disposal soft measurement method, an integrated neural network model is provided for measuring COD (chemical oxygen demand) of outlet water, BOD (biochemical oxygen demand) of the outlet water and TN (total nitrogen) of the outlet water, coupling relation between the three key water quality parameters is sufficiently utilized in the model, the integrated neural network model contains three feedforward neural sub-networks, and the various neural sub-networks are trained by particle swarm optimization, so that the optimal structure of each neural sub-network can be obtained. The COD of the outlet water, the BOD of the outlet water and the TN of the outlet water are predicted by the trained neural network finally, and prediction results are accurate.

A robot for gait training includes a walking-assist robot (100) configured to by put on legs of a walking trainee; a treadmill (200) with a conveyor belt floor which moves at a designated speed in order for the walking trainee to continuously perform gait training at a fixed position; a load-hoist (300) for upwardly supporting the body of the walking trainee; and a controller (400). The controller (400) includes an input unit (410) for receiving or inputting information or commands about size of the body of the walking trainee, and a speed, angle and rotational force of each joint required for training of the walking trainee, an information storage device for selectively storing the information and commands received through the input unit (410), a control unit for controlling a driving state of the walking-assist robot (100), the treadmill (200) and the load hoist (300) according to the information or commands input through the input unit (410) or transmitted from the information storage device, and a monitor (420) for numerically or graphically displaying the information transmitted from the walking-assist robot (100), the treadmill (200), the load hoist (300) and the information storage device. Therefore, it is possible to check the angle, speed and torque of each joint of the walking trainee in real time. As a result, by comparing the current walking of the walking trainee with a standard walking pattern appropriate for the training for the walking trainee, it is possible to analyze and determine whether the gait training is correctly performed and which walking pattern is more appropriate for the walking trainee.

The invention relates to a spinal orthotic device configured from one or more elements of a modular system, comprising the following elements:a lower abdominal corset (40, 120),an upper abdominal corset (17, 130) that can be attached cranially to the lower abdominal corset (40, 120),a corset supporting element (41) that can be secured posteriorly in the lower abdominal corset (40, 120) and is arranged along the lumbar spine, supporting the spine while restricting sagittal mobility,a thoracic spinal corset (10, 200) that can be attached cranially to the lower abdominal corset (40, 120),at least one curved supporting clasp (47) that can be inserted posteriorly optionally into a bandage of a lower abdominal corset (40, 120) and an upper abdominal corset (17, 130) or into an bandage of a lower abdominal corset (40, 120) and a thoracic spinal corset (10, 200), said curved supporting clasp being attached to a corset supporting element (41) for correction of lordosis and for restriction of sagittal and frontal mobility in the area of the lumbar spine,at least one supporting element (23, 160) which can optionally be secured cranially in the thoracic spinal corset (10, 200) and caudally to the corset supporting element (41, 150) and extends laterally along the spine to align and relieve the spine in the sagittal plane,and an abdominal truss pad (190) that can be attached ventrally to a lower abdominal corset (40, 120) for correction of lordosis of the lumbar spine and increasing the intra-abdominal pressure.

The invention discloses a graph convolutional neural network model and a vehicle trajectory prediction method using the same. The model is composed of an encoder module, a spatial information extraction layer module and a decoder module. The method comprises the following steps: firstly, sampling a predicted vehicle and surrounding vehicles in a traffic scene at a frequency of 5Hz, and collectingposition coordinates and kinetic parameters of each vehicle sampling point, including horizontal and longitudinal coordinates, horizontal and longitudinal vehicle speeds and accelerations; calculatingcollision time TTC between the predicted vehicle and surrounding vehicles according to the coordinates and speeds of the predicted vehicle and the surrounding vehicles, and judging vehicle behaviors;inputting each historical track of the vehicle containing the information into the model, encoding time sequence interaction features in the track, extracting spatial features, summarizing the features into context vectors, and inputting the context vectors into an LSTM decoder to generate future track coordinates of the vehicle. According to the method, the problem that feature information generated by vehicle interaction cannot be obtained by using a traditional recurrent neural network is solved, and the prediction precision of the vehicle trajectory is greatly improved.

A system, devices and methods for training of a health care worker, real time monitoring of a health care worker and providing real time corrective guidance for procedures related to hospital infection control. Furthermore, the devices, systems and methods of the present invention are configured to provide monitoring the performance and the quality of a hand hygiene procedure and providing real time, in accordance with the World Health Organization's “5 Moments Hand Hygiene” events protocol, corrective guidance to the health care worker procedures and practice related to hospital infection control.

The archery training device allows an archer to properly train his or her muscles, as well as train for proper archery technique. The archery training device includes a support having a plurality of pairs of apertures formed therethrough. A plurality of elastic members each has opposed first and second ends, which are respectively received within one of the plurality of pairs of apertures and attached to the support. Optionally, the upper end of an elastic band may be releasably attached to the support. A lower end of the optional elastic band is adapted for being held against a support surface by the foot of the archer. In use, the archer grasps the support with the bow hand and grasps a selected number of the plurality of elastic members with the string hand to simulate a string pull of an archery bow.

The embodiment of the invention discloses a risk model training and junk account detection method, device and equipment. The risk model training method comprises the following steps: acquiring an account network diagram, wherein the account network diagram includes nodes and edges connecting the nodes, the nodes represent accounts, and at least part of the edges represent that the accounts represented by the nodes connected with the at least part of the edges are registered through the same medium; determining a characteristic vector of the nodes through unsupervised learning; and training a risk model through supervised learning according to the determined characteristic vector. According to the risk model training method provided by the invention, a junk account is detected by using thetrained risk model, wherein the risk model is a GBDT (Gradient Boosting Decision Tree) classifier and the like.

The invention relates to an OCT (Optical Coherence Tomography) image layer segmentation method based on a neural network and a constraint graph search algorithm, and is designed for accurately segmenting a retina layer and new vessels. The OCT image layer segmentation method based on the neural network and the constraint graph search algorithm comprises the following steps that: obtaining OCT image features to train a neural network classifier; obtaining a final SF1 through a multi-resolution map search algorithm SF1; extracting 24 features of an OCT image, and using the neural network classifier to find initial surfaces S1, S2,...S8; according to initial boundaries S2 to S8, using the constraint graph search algorithm to find accurate SF2 to SF8 in sequence; and segmenting the new vessels and hydrops between SF7 and SF8. The OCT image layer segmentation method based on the neural network and the constraint graph search algorithm has the advantages of being simple in operation and accurate in detection results. The existing problems of low identification rate, poor segmentation effect and the like of a lesion OCT image segmentation algorithm.

An article of manufacture, more specifically, a basketball training glove, with a primary element that is a physical aid to train and reinforce proper ball handling technique, such as an element in the palm of the glove that makes it impossible for a ball handler to palm a basketball, enforcing the use of the finger tips for proper ball control, and another primary element of the glove with optional weights for resistance training for the purpose of building power and speed into the hands, wrists, and arms, the advantage of the current invention being that it provides for fingertip ball control and weight resistance training at the same time and incorporates both features into one pair of basketball training gloves.

The invention discloses a personal-credit evaluation method of intelligent combination and a system. The method is suitable for use in the system. The system includes a sample data acquisition module,a sample data processing module, a data feature training module, a scoring model construction module and a scoring model test module. The method includes the steps of: S1, acquiring sample data of model training by the sample data acquisition module; S2, dividing samples into a training set I, a training set II and a test set; S3, carrying out feature grouping on the sample data of the training set I, the training set II and the test set; S4, carrying out training on each set of data features, and generating a corresponding trained sub-model by each set of data features; S5, carrying out prediction on the trained sub-models; and S6, testing a finally constructed credit scoring model on the test set by the scoring model test module to evaluate model test effect.

The invention provides an operation cutting training system and an operation cutting training method based on force feedback and used for a surgical robot, and relates to an operation cutting training system used for the surgical robot, in particular to the operation cutting training system based on force feedback. The invention aims at solving the problem that the real-time property of a finite element model structure in the existing operation cutting training system used for the robot is poor. The system comprises a 3d virtual environment and 3d surgical instrument model building module for building a 3d virtual environment and a 3d surgical instrument model, a 3d virtual soft tissue model building module for building a 3d virtual soft tissue model, a model reading and positioning module for loading the size, position and rendering modes of the model, a cutting tool and model collision detection module for determining the collision position and time, a force feedback module for realizing force touch and completing the force feedback operation, and a classification cutting module for realizing face cutting and volume cutting. The system and the method provided by the invention are suitable for the operation cutting training of the surgical robot.

The invention provides a multi-style traditional Chinese painting peanut forming method based on a neural network, and the method comprises the steps: mapping from white painting to crafts or Chineseink painting through employing an attention architecture, carrying out training through employing a non-paired data set to acquire crafts and Chinese ink paintings with certain style characteristics.According to the invention, a new loss MSSSIM loss function is introduced into a GAN-based attention architecture; the distance between the reconstructed image and the source image is compared in a multi-scale manner from three aspects of brightness, contrast and structure, and the image style and content are guided to be accurately generated together with a cyclic consistency loss L1 function, sothat the problems of unclear original network picture, artifacts, fuzzy ink marks, noisy points and the like are solved. According to the method, the problem that high-quality pictures with clear pictures cannot be generated due to abstraction and complexity of traditional Chinese painting styles in research of a traditional method is solved.

The invention discloses a lesion monitoring method and device, a computer device and a storage medium. The lesion monitoring method comprises the following steps: inputting the sample data of an original CT image into a preset segmentation model for segmentation calculation, and outputting segmented liver image data; inputting the sample data of the original CT image and the segmented liver imagedata into a preset identification model for calculation, and outputting an identification result, wherein the segmentation model and the recognition model are respectively trained through a first convolution neural network and a second convolution neural network, and the first convolution neural network and the second convolution neural network are in cascade arrangement. According to the method,the neural network is used for learning liver features and lesion features in the original CT image, and the relation between the CT slices and the tags is established through the two cascaded full-convolution neural networks to distribute the task training model, so that the optimal network parameters can be found as soon as possible, and the disease treatment efficiency and the accuracy of disease analysis are improved.

The invention provides a vegetable identification method and device. The method includes the steps of establishing the first convolution neural network model, collecting the images of each vegetable in different scenes, building the corresponding input dataset, iteratively training the first convolution neural network model according to a learning rate adaptively updated gradient descent algorithmand the input data set, obtaining the second convolution neural network model, carrying out the vegetable recognition according to the second convolution neural network model, and improving the efficiency of the vegetable recognition is improved. The first convolution neural network model is trained by a stochastic gradient descent algorithm, so that the second convolution neural network model obtained by the training can accurately identify the vegetable species under the complex background.

The invention discloses a handwritten sample recognition method and system based on sample enhancement, and the method comprises the steps: S1, generating a labeled sample, marking handwritten characters in an image sample, cutting the handwritten characters out of the image sample, and carrying out the classification of the cut handwritten characters; S2, sample enhancement: carrying out random transformation on the labeled samples to generate transformed samples, and generating enhanced samples in the same distribution as the transformed samples by utilizing a generation model; S3, sample synthesis: generating a training sample by using the enhanced sample; S4, model training: a training sample is used to train a detection classification model and a handwritten sample identification model; and S5, identification application: detecting the position of the handwritten character by using the trained detection classification model, and then identifying the handwritten character by usingthe handwritten sample identification model. The detection classification model and the recognition model are optimized by increasing the diversity of the training samples, and the problems that in the prior art, the offline handwriting recognition accuracy is low, the handwriting recognition samples are difficult to mark, and the model is lifted slowly are effectively solved.

The invention relates to an improved Lasso+RBF neural network combined prediction model, and belongs to the field of big data analysis and processing. A prediction process of the model comprises the following steps of: defining life cycle features of a customer relationship so as to divide a customer life cycle into an obtaining stage, a lifting stage, a mature stage, a decline stage and a loss stage; taking customers in the loss stage as a training set and a test set of the model, taking customers in the other four stages as prediction customers, and subdividing the customers in the loss stage into the former four stages; respectively extracting features by using Lasso regression and respectively training an RBF neural network model corresponding to each stage; respectively substituting the customers in the former four non-loss stages into the trained models corresponding to the stages to carry out prediction; and finally combining the obtained prediction results to obtain a to-be-lost customer set. According to the method provided by the invention, the extracted features are more correct, the unbalance of data is decreased and the prediction accuracy is improved.

A breath testing apparatus is provided with on-board “Instruction Manual” tutorials incorporated in a display device installed in the device. Alternatively, on-board instructional videos can be added to various objects or processes requiring instructions.

The invention relates to the field of artificial intelligence. The embodiment of the invention provides a method and device for detecting a target object in an image. The method comprises the following steps: executing the following prediction operations by using a pre-trained neural network: detecting a target object in a two-dimensional image, and determining a two-dimensional bounding box of the target object; determining a relative position constraint relationship between the two-dimensional bounding box of the target object and a three-dimensional projection bounding box obtained by projecting the three-dimensional bounding box of the target object into the two-dimensional image; and determining a three-dimensional projection bounding box of the target object according to the two-dimensional bounding box of the target object and the relative position constraint relationship between the two-dimensional bounding box and the three-dimensional projection bounding box of the target object. According to the method, the accuracy of target object position detection is improved.

The invention discloses a generative adversarial network model and a vehicle trajectory prediction method using the generative adversarial network model, the generative adversarial network model is composed of two modules: a first module: a generator of a generative adversarial network; and a second module: a discriminator of the generative adversarial network. The generator is used for generatinga simulation track, the discriminator is used for discriminating the authenticity of the track, the first module and the second module are in game confrontation with each other, and finally the generator capable of generating the simulation track highly similar to the real track is obtained. The generative adversarial network model is trained, and a data acquisition method, an optimizer, a learning rate, a training round, an iteration round and a loss function are given; finally, a method for verifying the accuracy of the trajectory generated by the vehicle trajectory prediction system is provided. According to the prediction method for track prediction by using the generative adversarial network provided by the invention, the past network structure is innovated, and the prediction accuracy is effectively improved.

A method for manufacturing a three-dimensional anatomical structure comprising the steps of obtaining a plurality of medical images containing at least one anatomical region of interest, wherein each anatomical region of interest is defined based on a grey level value on the medical images; segmenting each medical image based on the anatomical region of interest to obtain the grey level value; converting the grey level values of the respective segmented medical images into vector data; interpolating the vector data of each segmented medical image to form a three-dimensional data; and producing a three-dimensional anatomical structure of the region of interest from the three-dimensional data.

The invention discloses a deformable convolution kernel method based on a WGAN (Wasserstein-Generative Adversarial Network) model, and belongs to the field of the deep learning neural network. The method comprises the following steps that: S1: constructing an original generative adversarial network model; S2: constructing a Wasserstein distance as the judgment index of the adversarial network model; S3: initializing random noise, and inputting the random noise into a generator; S4: in the WGAN model, utilizing a deformable convolution kernel to carry out convolution on an image; and S5: inputting a loss function obtained by a deformable convolution operation into the generator for subsequent training. By use of the deformable convolution kernel method, which is constructed by the invention, based on the WGAN model, a convolution way generated after a discriminator and the generator receive a picture is changed, the discriminator and the generator can automatically change the size of the convolution kernel according to a training situation, so that the characteristics of a dataset image can be adaptively learnt, and the robustness of whole network training is improved.

The invention discloses a touch-screen multifunctional non-human primate animal cognitive-function testing cage and belongs to the technical field of animal experiment devices. The touch-screen multifunctional non-human primate animal cognitive-function testing cage comprises a testing cage and a transfer cage. An infrared-induction touch screen connected with a computer is fastened on the rear vertical face of the testing cage capable of moving back and forth, an observation window and a movable camera are arranged on the top face, a water supply nozzle is located at the rear locking-fixing position capable of moving vertically in front of the touch face of the touch screen, a feces receiving plate which can be pushed into or pulled out of the lateral vertical face of the testing cage is arranged under a grid-shaped base, and a testing cage door which can be pushed into or pulled out of the side face of the testing cage is arranged on a grid-shaped front vertical face. The transfer cage is provided with truckles and an observation window, and the front vertical face is a push-pull type transfer cage door which can be in butt joint with the testing cage door after being opened. The touch-screen multifunctional non-human primate animal cognitive-function testing cage can be used for flexibly, conveniently, intelligently, efficiently and accurately performing non-human primate animal training and testing based on a modern computer technology, does not need an additionally-equipped corresponding cage, has multiple practical functions and is low in purchase cost and usage cost.

The invention discloses a rumor detection method combining a self-attention mechanism and a generative adversarial network. The rumor detection method comprises the steps of collecting rumor text datato form a rumor data set; based on a self-attention mechanism, constructing a generative adversarial network generator comprising a self-attention layer; constructing a discriminator network, and respectively carrying out rumor detection and classification on the original rumor text and the text decoded by the generator; training the generative adversarial network, and adjusting model parametersof a generator and model parameters of a discriminator; and extracting a discriminator network of the generative adversarial network, and performing rumor detection on the to-be-detected text. Compared with an existing rumor detection method, the rumor detection method is higher in detection precision and better in robustness; a self-attention layer is adopted in the generator, key features are constructed through semantic learning of rumor samples, text examples rich in expression features are generated to simulate information loss and confusion in the rumor propagation process, and the semantic feature recognition capacity of the discriminator is enhanced through adversarial training.

The invention provides an image processing method and an image processing device. The method comprises the following steps of: constructing a map environment based on map data; recording the video picture of the vehicle during driving and the driving data of the vehicle by a video camera, wherein the video picture includes: a video picture of a peripheral vehicle and a traffic environment change during driving of the vehicle; according to the video picture and the driving data of the vehicle, simulating the driving state and the traffic environment change of the surrounding vehicles in the mapenvironment, and constructing the test scene, extracting the image picture of the test scene and the information of the object unit in the image picture of the test scene, and labeling the image picture according to the information of the object unit; saving an image picture of the annotated test scene as a training material for perceptual training of the autopilot system. The present applicationeliminates the huge cost of purchasing a large amount of image data and labeling pictures manually, improves labeling efficiency and accuracy, and makes the trained neural network more accurate.

The embodiment of the invention provides a machine translation model training method and device and electronic equipment. The machine translation model training method comprises the steps: employing afirst parallel corpus, and carrying out preliminary training on a translation model from a target end to a source end; utilizing the trained translation model from the target end to the source end totranslate each sentence in a definite monolingual corpus to obtain a synthetic corpus, and splicing the synthetic corpus with the definite monolingual corpus to obtain a second parallel corpus; evaluating the credibility of a translation result obtained by using the trained translation model from the target end to the source end by using a Monte-Carlo random inactivation algorithm; and based on the credibility, training a translation model from a source end to a target end by utilizing an integral corpus formed by the first parallel corpus and the second parallel corpus. According to the embodiment of the invention, the translation model can still be accurately trained in the presence of noise interference, and the accuracy of the translation model is ensured.

A robot for gait training includes a walking-assist robot (100) to be put on legs of a walking trainee; a treadmill (200; a load-hoist (300) for supporting the body of the walking trainee; and a controller (400). The controller (400) includes an input unit (410) for receiving or inputting information or commands, and a speed, angle and rotational force of each joint required for training of the walking trainee, an information storage device for selectively storing the information and commands received through the input unit (410), a control unit for controlling the walking-assist robot (100), the treadmill (200) and the load hoist (300) according to the information or commands input through the input unit (410) or transmitted from the information storage device, and a monitor (420) for displaying the information transmitted from the walking-assist robot (100), the treadmill (200), the load hoist (300) and the information storage device.

The invention relates to the technical field of user portraying, and provides a federation model training method and device, a customer portraying method and device, equipment and a medium, and the method comprises the steps: obtaining a participant list and an initial customer portraying federation model, and screening out qualified participants from the participant list according to a preset screening scheme; sending the initial customer portrait federation model to each qualified participant; receiving returned model parameters; performing abnormal feature extraction through a malicious parameter detection model by applying an MPI parallel method, and outputting an identification result of each model parameter according to the extracted abnormal features; filtering malicious parameters to obtain final normal parameters; and performing updating and federation learning to obtain a global customer portrait federation model. According to the method, the MPI parallel method is applied, abnormal feature extraction and malicious parameter filtering processing are carried out through the malicious parameter detection model, malicious parameters provided by malicious participants are automatically removed, and the efficiency and precision of federal learning modeling are improved.

The invention discloses a geographical name speech signal recognition method based on continuous Gaussian mixture HMM models (hidden Markov model), wherein a process of training the continuous Gaussian mixture HMM model comprises the following steps: defining and initializing the HMM models; putting a characteristic matrix of a category of geographical name speech signals into the models and conducting training; calculating a probability that the category of geographical name speech signals appear in accordance with model parameters; comparing the probability with an output probability before the training is conducted, and judging whether a relative error satisfies output conditions or not; if so, outputting the HMM model corresponding to the category of geographical name speech signals; if not, judging whether training times reach a maximum training threshold or not; conducting the training once again when the training times fail to reach the maximum training threshold, or outputting the HMM models when the training times reach the maximum training threshold; and putting characteristic matrices of the various categories of geographical name speech signals into the models so as to obtain a plurality of HMM models corresponding to different geographical names, so that a geographical name speech recognition model library is formed. With the application of the geographical name speech signal recognition method provided by the invention, the HMM models, which are applicable to geographical name speech recognition of isolate words, as well as the geographical name speech recognition model library can be obtained, so that conditions are created for conducting geographical name speech recognition accurately.