56 results about "Mental image" patented technology

The invention relates to a three dimensional (3D) virtual fitting method based on somatosensory technology. The 3D virtual fitting method based on the somatosensory technology is characterized by including the steps: step one, setting up a hardware system; step two, tracking human skeleton; step three, reconstructing three dimensions; step four, conducting real-time measurement on three dimensional sign information of a human three-dimensional model, utilizing a measured characteristic size to adjust the size of corresponding place of a garment model and enabling the garment model to fit in with a human body figure and to be superimposed on the human three-dimensional model; steps five, fitting in a scene. Due to the novel information technology of somatosensory location, strengthening reality and image recognition, the 3D virtual fitting method based on somatosensory technology shoots and manufactures real garments into high definition images which fit in with a certain specification, utilizes a depth camera sensor to collect image information of human skeleton and depth, perfectly mixes the garments and real images and produces display effects of different garments in different scenes in a mode of strengthening reality.

The invention discloses a face image restoration method based on a multi-discriminator generative adversarial network, and the method comprises the following steps: (1) carrying out the preprocessingof images in a disclosed face image database, and inputting the images into a generator, so as to obtain a generated image; (2) inputting the real image and the generated image into a plurality of discriminators to obtain a feedback value; (3) taking the feedback values of the plurality of discriminators as confrontation loss, and performing confrontation training on the generative adversarial network by combining the perception loss and the reconstruction loss; and (4) inputting the missing face image into the trained generator to obtain a repaired face image. Aiming at the problem of repairing the shielded or damaged face image, a generative adversarial network structure with multiple discriminators is adopted, so that the problem of low authenticity of the repaired image is solved, andthe repaired image is more natural and more real.

The invention discloses a multi-image face alignment method based on the generative adversarial network. The method comprises that multiple real face images are input to a generator, and the generatorprocesses the images to a synthesis image which fits distribution of the real images; the real face images and the synthesis image are input to a discriminator to obtain the real probability of the real image and the real probability of the synthesis image, and parameters of the generator and the discriminator are updated iteratively till the parameters are converged so as to determine a model established by the generator and the discriminator; and the face images to be aligned to input to the determined model, and an aligned face image is obtained by once forward transmission operation. Thus, one clear and aligned face image can be generated according to multiple face images, the sparse degree of a noise matrix obtained by subtracting the aligned image from a multi-face-image matrix reflects the degree in which the generated face image is aligned with the input face images, and the generative adversarial network records integral and detailed characteristics of the face.

The invention provides a multi-view and multi-state gait recognition method. The method comprises the following steps of: carrying out contour extraction and time synchronization on a gait video in a training set; establishing projection relations among presentative expected value, visual angle, state and identity; carrying out visual angle estimation on a human body gait contour sequence of a prototype set so as to obtain identity information from the projection relations among the presentative expected value, the visual angle, the state and the identity; carrying out visual angle estimationon a human body gait contour sequence of a testing set so as to obtain identity information from the projection relations among the presentative expected value, the visual angle, the state and the identity in the training set; and comparing the identity information with all existing identity vectors in the prototype set so as to judge the person in the prototype set. The multi-view and multi-state gait recognition method provided by the invention is different from the traditional method based on the whole gait period that the variation process of the gait motion is more accurately modeled by using single state as minimum unit, multi-state model under each visual angle is trained and the gait visual angle of the video to be identified can be estimated.

The invention discloses a method and a system for continuous attention assessment. The method comprises the following steps: when a tested person is in a simulated task state, the behavior index data, the eye movement index data and the electroencephalogram index data of the tested person are acquired; the above acquired data are input into a preset comprehensive evaluation model, and the output of the comprehensive evaluation model is adopted as one of a plurality of assessment grades. The comprehensive evaluation model is obtained in advance through the following mode: an ideal evaluation index table of the plurality of assessment grades is obtained based on the test data of tested people of a preset number in the actual task state, and a multi-input and single-output relation model is obtained through the neural network learning process under the supervision of the ideal evaluation index table. According to the invention, the continuous attention is fully fused with specific scene task characteristics, and indexes of three dimensions are adopted. A more objective and more accurate continuous attention assessment scheme is established from the representation layer to the mechanism layer, and from exogenous factors to endogenous factors.

The invention relates to a method for detecting a plurality of pedestrians in a crowding scene. The method comprises the following steps: a, inputting a training set containing a plurality training sample images, and dividing pedestrians in the training sample images into a plurality of subclasses, wherein each subclass is corresponding to one pedestrian change; b, establishing a training set pedestrian representation feature model; c, establishing a spatial relational model of pedestrians on the training set; d, establishing a probabilistic model to describe the pedestrians in the crowding scene, and converting the probabilistic model into an energy objective function; e, looking for the optimal parameter of the energy objective function in a parameter learning method based on a latent rank SVM (Support Vector Machine) so as to obtain a determined energy objective function; and f, inputting crowding scene pedestrian images to be detected, and detecting by a model deduction method based on expansion move and the determined energy objective function to obtain the result of the crowding scene pedestrian images to be detected. According to the method disclosed by the invention, the detection precision is improved, and the application range of the method is wide.

A method for playing computer games on a computerized game system having a display, memory and one or more input devices. The method includes one or more acts of pre-game activity and one or more acts of game activity. The pre-game activity includes learning rules of the game including observing one or more objects related to the game on the display, thereby forming a visual percept of the objects, and transforming the visual percept of the objects into a mental image percept of the objects, in the absence of a visual percept of the objects. At the game activity one or more of objects are invisible. The game activity includes activating one or more of the input devices and inputting data in dependence upon the mental imagery percept of the invisible objects, thereby forming play input data. The game activity further includes receiving feedback that corresponds to the play input data and activating one or more of the input devices to form additional play input data in dependence upon the received feedback.

The invention discloses a clothing push method and a system based on an intelligent television. The method comprises steps that: when users use the intelligent television, image information of the users is acquired and television contents currently played on the intelligent television are obtained; according to the acquired image information and the television contents currently played, identities of the users are analyzed and mental ages of the users are obtained, and the users are classified according the metal ages of the users; and clothing types matching classification results of the users are selected according to the classification results of the users, and the matching clothing is pushed to the users according to the clothing types.

The invention discloses a vision-based intelligent vehicle transverse control method. The method comprises the following steps: step one, a camera acquiring an image; step two, according to the image acquired by the camera in acquiring a front vehicle image in real time, detecting whether a front vehicle exists by use of a computer vision technology; step three, carrying out vehicle longitudinal and transverse distance measurement; step four, carrying out vehicle angle deviation calculation; and step five, performing steering wheel angle control. The camera acquires the front vehicle image in real time in the first step, image analyzing is performed on an acquired initial image, then exposure parameters of the camera are adjusted according to an image analyzing result, whether the picture depicts a daytime scene or a night scene is determined according to the image acquired by the camera, and different exposure strategies are employed for exposure according to different scenes so that vehicle detection requirements for different scenes are met.

Systems for matching face shapes may include a computer-readable non-transitory storage medium and an executing hardware unit. The storage medium may include a set of instructions for target object shape matching. The executing hardware unit may be in communication with the storage medium and may be configured to execute the set of instructions. The executing hardware unit may be configured to obtain a target object image for shape matching; determine a shape character of the target object image based on a shape of the target object image; determine similarities between the target object image and a plurality of template images of reference objects based on the shape character of the target object image and shape characters of the reference objects in the plurality of template images; and select a template image from the plurality of template images that has a largest similarity to the target object image.

The invention discloses a three-dimensional face age classifying device and method based on three-dimensional point cloud. The device comprises a characteristic region detection unit for positioning a characteristic region of the three-dimensional point cloud, a mapping unit for mapping the three-dimensional point cloud into depth image space, a characteristic calculating unit for calculating the depth image presentation characteristics of a mapped depth image and an age classifier calculating unit for carrying out age classifying based on the depth image presentation characteristic. The method includes the steps of detecting the characteristic region, mapping the depth image, calculating the image presentation characteristic, and carrying out classifying. According to the device and method, the mode that an image presentation characteristic pool is built in cooperation with a plurality of textural characteristics is used, the characteristics of a three-dimensional depth face image are accurately described, then accurate classifying is achieved through an age random forest classifier on the basis of an image presentation characteristic set, and the classifying accuracy is high.

The invention discloses a personalized garment recommending method based on garment image and label text double-mode content analysis. The personalized garment recommending method comprises followingsteps: S1: by analyzing detailed properties in garment images on a shopping website, establishing a garment product model and a user preference model which are described in the form of images and features of garment detailed parts; S2: by analyzing garment label texts on the shopping website, establishing a garment product model and a user preference model which are described in the text form; andS3:combining the garment product model based on the garment images and established in the step S1 with the user preference model established in the step S2 to generate a recommending result. According to the invention, image information and text information of garments are combined, by fusing the above models, personalized garments can be recommended to the user.

The invention discloses an intelligent homologous health treating and preventing medical mode. Through establishing an accurate sign, behavior and symptom module and corresponding case module connecting system, problems of non-obvious symptom phenomenon, difficult targeted diagnosis and medical industry difficulties in health treating and preventative treating. A health treating operation mode which is similar with four steps of symptom, case, targeted diagnosis and treatment effect of a disease treating mode is established. The health treating mode can be connected with organizations of hospitals, enterprises and public institutions, community, home and each individual, a health ecological system which comprises signs, behaviors, health and diseases is formed. The mode improves physical agility, life quality and working capability of the user and further improves human resource competition capability of organizations. Furthermore the mode improves development of a disease medical system and facilitates a doctor in quickly searching an accurate disease treatment method.

The invention discloses a shopping image collaborative segmenting method for performing collaborative segmentation on a plurality of input shopping images containing one piece of garment automatically and extracting garment areas in the images. According to the invention, foreground (garment) areas and background areas in the shopping images are positioned by using an upper body detection algorithm and an image distinctive area detecting algorithm; a Gaussianmixed model is adopted for performing unified modeling on the foreground and the background of all the garment images and a collaborative segmentation objective function is established; then a minimum cut algorithm is used for performing iteration solving of the objective function; and finally the segmentation results of all the images are obtained. The method is suitable for being applied to e-commerce websites and online shopping websites. Interference on image retrieval caused by complex background can be eliminated and searching accuracy of searching engines can be improved.

In an image of lithiasis in a computed tomography examination, a segment of the image is chosen with a detection of the maximum of the gradient of radiological density measured in the voxels of the reconstructed image. This mode of action gives an evaluation of the envelope of the lithiasis that is objective and independent of the practitioner who performs it. Furthermore, to make the characterization of the lithiasis more realistic, especially to follow the effects of a therapeutic treatment, a characteristic result is given, equivalent to the volume of the lithiasis, preferably a weighted volume, and to the mean density of the lithiasis. The disclosed method also can facilitate the scheduling of surgical operations when it is not possible to administer oral treatment or carry out lithotripsy.

A system for generating simultaneous awareness of two separate sensor stimulation patterns (1,5,15,16, 23L,23R,24L,24R) and implementing interest measurement (20,22,7) and attention factors modification (4,10,17,18) in different left and right stimulation patterns in order to suppress stereo vision and binocular rivalry and instead facilitate simultaneous observance of both mental images (9). The user will be able to use the balance point between left and right image rivalry to maintain awareness of a simultaneous dual vision, audio, and tactile experience. Such a doubling experience can be used to:extract information and make decisions on two simultaneously observed world views.maintain two simultaneous views of different computer screens to increase performance and reduce fatigue.encourage the fusion of two different simultaneous experiences into an abstract information synthesis instead of the conventional three dimensional world view.perform measurements on interest an operator has on a left right scene content.

The invention discloses a target tracking method on the basis of online study for a context relation of a space-time structure. The method is used for tracking target images in videos, and includes steps of creating a time-structure context model and a space-structure context model for target tracking; sampling areas, where an object probably appears, in a video frame to obtain sample images; inputting the sample images into the time-structure context model and the space-structure context model to obtain time probability results and space probability results which indicate various sample points in the sample images are areas with the target; and fusing the space probability results and the time probability results to obtain posterior probabilities indicating that the various sample points belong to the areas with the target, and using a sample point corresponding to the maximum value of the posterior probabilities as a final result of target tracking. The target tracking method has the advantage that the target can be accurately positioned via automatic online study for representation features of the target under the condition that any information of the target and any information of scenes with the target are unknown.

The invention discloses a head-mounted VR vision improving system based on binocular vision and mental image training. The system comprises a head-mounted VR vision improving training device and a client APP. The head-mounted VR vision improving training device comprises a binocular camera and a VR imaging module. The binocular camera is disposed directly in front of a casing. The VR imaging module comprises a microprocessor, a wireless communication module, a micro motor and a binocular lens. A binocular camera, a brain wave monitoring device and a brightness sensor are respectively connectedto the microprocessor. The microprocessor communicates with the client APP through the wireless communication module. The novel vision improving training system has high security, low cost and high intelligence degree and is of great significance in the field of refractive error treatment.

The invention discloses a region-of-interest medical image chaotic encryption method based on an information entropy update key, and relates to the technical field of image information security, whichcan overcome the security defects that an existing medical image encryption technology is low in encryption efficiency and cannot resist attacks of known plaintexts and selected plaintexts. Accordingto the information entropy updating key-based region-of-interest medical image chaotic encryption and decryption method provided by the invention, the region-of-interest of the medical image is processed by utilizing wavelet transform, so that the frequency domain encryption of the image is realized. Meanwhile, the hyperchaotic system is used for scrambling and diffusing the region of interest ofthe medical image, so that the protection of the image in a spatial domain is realized. Besides, according to the information entropy key updating method designed by the invention, the key is not only related to the parameters of the chaotic system, but also related to the plaintext image, so that known plaintext attacks or selected plaintext attacks can be resisted.

The invention discloses a multimedia information processing method and device. The method comprises the following steps: obtaining presentation information, wherein the presentation information comprises electromagnetic field spectral information for showing an object and capable of being observed by naked eyes and / or capable of being collected by a device; establishing a four-dimensional space-time model for showing the presentation information according to the obtained presentation information, wherein the four-dimensional space-time model has attributes capable of showing the time-dependent presentation information in a digital manner; and carrying out coding processing on the established four-dimensional space-time model, and sending the four-dimensional space-time model obtained after coding processing. In the scheme, the established four-dimensional space-time model has the attributes capable of showing the time-dependent presentation information in a digital manner, so that the problem of time delay is solved to some extent. The method and device solve the defect of time delay in the prior art to some extent.

The invention discloses an image feature decomposition-based high spatial resolution remote sensing image prospecting method. The method comprises the following steps of extracting color feature and the texture feature of an image; solving the spectrum second derivative of the image; comprehensively analyzing the three elements i.e. the extracted color feature, the texture feature and the second derivative image; and by using a logic overlay analysis method, solving the intersection by using the algorithm of multiplying the three elements through a fuzzy mathematical theory, wherein the intersection position is the remote sensing prospecting target area. According to the method, appearance shapes such as color, shape, texture and the like shown by the high spatial resolution remote sensing data are fully utilized and the data is fully mined by adopting the techniques such as mathematics, image processing, GIS (Geographic Information System) and the like to comprehensively confirm the prospecting target area; and reliable basis is provided for the periphery prospecting of mining areas and the ore deposit location in the next step.

The invention discloses a method and a device for presenting multimedia information. The method comprises the following steps: receiving a four-dimensional spatial-temporal model used for presenting representation information, wherein the four-dimensional spatial-temporal model has an attribute which can present the representation information along with time changes in a digital form; carrying out decoding processing on the four-dimensional spatial-temporal model to obtain a decoded four-dimensional spatial-temporal model; and according to the decoded four-dimensional spatial-temporal model, presenting the representation information presented by the four-dimensional spatial-temporal model. In the scheme, the four-dimensional spatial-temporal model has the attribute which can present the representation information along with time changes in a digital form, so that the problem of time delay of the presented representation information is solved to a certain degree, and therefore, the scheme solves the defect of the time delay in the prior art to a certain degree.

Methods, systems and computer program products for retrieving mental images of faces from the human brain. An exemplary embodiment includes a method for retrieving mental images of faces from a human brain, the method including generating an image on a screen, measuring brain activity data from a brain scanner, varying characteristics of the image on the screen, detecting changes in the brain activity data measured from the brain scanner, and generating a final image on the screen, the final image being related to a subjectively described mental image.

The invention discloses a method for estimating perceptual semantic content by analysis of brain activity. Provided is an estimation method for measuring and analyzing brain activity to estimate a perceptual semantic content thereof. This method comprises: (1) inputting, to a data processing means, an output when a cranial nerve activity detection means detects an annotation of a perceptual content and brain activity induced in a subject by a training stimulation; (2) associating a sematic space representation of the training stimulation and the output of the cranial nerve activity detection means in a stored semantic space and storing the association in a training result information storage means; (3) inputting, to the data processing means, an output when the cranial nerve activity detection means detects brain activity induced by a novel stimulation, and obtaining a probability distribution in the semantic space which represents perceptual semantic contents for the output of the novel stimulation-induced brain activity by the cranial nerve activity detection means on the basis of the association; and (4) estimating a highly probable perceptual semantic content on the basis of the probability distribution. The association process may be performed for each subject. In the probability estimation process, the likelihood calculated on the basis of the coordinate of a given word in the semantic space and the probability distribution is used as an indicator.

The invention discloses a method and system for assessing the update ability of spatial working memory. The method includes the steps of: obtaining the behavior index data and EEG index data of the tested person when the tested person is in a simulated task state; The data is used as the input of the preset comprehensive evaluation model, and the output of the comprehensive evaluation model is one of multiple assessment levels; the comprehensive evaluation model is obtained in advance in the following way: based on the test data of a predetermined number of tested personnel in the actual task state, the obtained An ideal evaluation index table corresponding to the assessment level, and under the supervision of the ideal evaluation index table, a multi-input and single-output relationship model is obtained through neural network learning. The implementation of the present invention fully integrates the updating ability of spatial working memory with the characteristics of specific scene tasks, adopts two-dimensional indicators, goes deep from the appearance level to the mechanism level, and from exogenous factors to endogenous factors, establishes a more objective and accurate Spatial working memory update ability assessment scheme.

The invention discloses an image-text double-coding mechanism implementation model based on a CR2 neural network, and relates to the field of human cognition and knowledge representation. The image-text double-coding mechanism implementation model based on the CR<2> neural network is characterized in that the implementation model includes a representation system which adopts a multilayer Convolutional Neural Network (CNN) model to obtain ''image units'' representing mental images of information; and a semantic system which adopts an RNNLM language model to obtain ''language units'' representing information semantics; a reference association system which adopts an RBF self-propagation neural network model, wherein a positive model uses the ''image units'' as input, output is the ''language units'' associated for reference, and an inverse model is inverse operation of the positive model. Here the CR<2> neural network refers to an organic composition of three neural networks of CNN, RNN and RBF. The image-text double-coding mechanism implementation model based on the CR2 neural network realizes establishment of models of an image representation system and a natural language semantic system, and also establishes a model of reference association between the two systems. The models completely simulate the whole process of an image-text double-coding cognition mechanism.

The embodiments of the present invention provide an object detection method, an object detection device, and an electronic apparatus. The object detection method includes the following steps that: a plurality of super pixels in an image to be detected are acquired; the representation features of each super pixel and spatial context information between the plurality of super pixels are extracted; and the category and / or location of at least one object included in the image are / is determined according to the representation features and the spatial context information. With the object detection method and the object detection device of the embodiments of the invention adopted, object detection is carried out based on super pixels, and therefore, objects missed by a traditional candidate framegeneration algorithm can be naturally detected out, and the accuracy of object detection can be effectively improved.

The invention belongs to the technical field of road traffic signs, and discloses an urban road intersection traffic sign line and a positioning method thereof to direct vehicles. The urban road intersection traffic sign line has the characteristic of being continuous, facilitates improvement of spatial cognitive psychology of divers, facilitates correction of wrong visual image and orientation concept of people, facilitates reduction of mental fatigue due to anxiety and confusion caused by path selection and judgment, and facilitates formation of longtime effective memory and judgment. According to the positioning method, only simple parameters need to be conveniently measured to determine the front distance of traffic signs, and further sign board color indicator bars and road colored guide lines can be arranged to lead the divers to reasonably park based on driving needs. Urban traffic management departments can erect and spray traffic sing lines with the best management effect according to the method. The colored sign lines are simple in principle, low in construction difficulty, high in speed, low in project investment, and is suitable for popularization and application.

The invention relates to a model of a rat suffered from heat blood stasis syndrome, which makes the induction by continuous LPS abdominal cavity injection of the rat. The model is characterized in that the continuous LPS abdominal cavity injection is made twice a week, namely one injection at intervals of three days or four days. The rat model makes the induction by continuous LPS abdominal cavity injection of the rat, so that the established heat blood stasis syndrome animal model is good in stability and long in survival time, and the simulated symptoms are nearly line with the actual clinical performance and the objective detection indexes, not only the model is more similar to the presentation of the blood stasis syndrome in extrinsic feature, but also the mode can be judged by utilizing a plurality of more detailed biological indicators.