45 results about "Normal face" patented technology

The invention discloses a face occluder identification method. The method comprises the following steps of: acquiring a plurality of images from a video camera system, and determining the region wherein the face is positioned in each image of the plurality of images; counting coordinate values of the central position of each organ and the number of the coordinate values in the plurality of images having the face image region; determining the coordinate value, in accordance with the central position of each organ in a normal face image, of the coordinate values as the position of an organ; and determining the occlusion situation of the position of the organ according to the number of the coordinate values of a certain organ, and determining the abnormal characteristics of the position of each organ in the face image. According to another aspect, the invention also discloses a face occluder identification device. The face occluder identification device is used for acquiring image data from the video camera system to determine the abnormal characteristics of the position of each organ in the face image. By the method and the device, abnormal situations such as an occluder and the like at the positions of main organs of the face can be identified efficiently, and the method and the device are particularly applicable for video systems, needing vigilance and surveillance, of banks, jewelry shop and the like.

The invention provides a fake face detecting method and a system for realizing same, wherein the fake face detecting method and the system belong to the technical field of video image identification. The fake face detecting method comprises the following steps of acquiring a person entering signal, determining whether a person enters, and if yes, performing pretreatment on a video image; detecting a face for judging whether the face or head exists; if yes, detecting a nose for judging whether the nose exists; if no, judging as a fake face and performing alarming; if yes, estimating detecting areas of eyes and mouth, then detecting the eyes and the mouth for judging whether the eyes and the mouth coexist; if yes, judging the face as a normal face; and otherwise, judging the face as the fake face and performing alarming. The fake face detecting method and the system have beneficial effects of performing real-time detecting and alarming on the fake face of an ATM, utilizing the nose as guidance for detecting the eyes and the mouth, and realizing higher accuracy in positioning organs on the face.

The invention belongs to the technical field of video frequency image analysis and particularly discloses an abnormal face detecting method for embedded DVR system. In the invention, face detection and background identification technology is applied in a digital signal processor (DSP) processing step of the embedded DVR so as to determine that whether people exist in a current image and whether facial features are clear and not sheltered, and if abnormal facial features are detected, an alarm signal is sent to remind related personnel to pay more attention to the current conditions. When the method of the invention is adopted, the technology of detecting whether normal faces exit in a scene is achieved simply by a mode of image processing, and alarm effect that lawbreakers are prevented from committing a crime at an ATM machine of self-service banking by shielding faces is achieved when the method is used in combination with the embedded DVR.

A packet communication apparatus that allows the user of the communication terminal performing packet communication to take countermeasures when packet loss occurs, such as, moving to a more satisfactory communication environment is provided. A mark M3 indicating packet loss are displayed on the display screen 119 of the apparatus in addition to marks M1 and M2 indicating the received power and the remaining amount of the battery, respectively. The packet loss is notified to the user by changing the expression of a face of the mark M3 according to the extent of the packet loss. That is, when the mark M3 has a smiling face, a state in which the packet loss is not occurring is indicated, when the mark M3 has a normal face, a state in which the packet loss has occurred but has substantially no practical problems as the packet is compensated is indicated, and when the mark M3 has a crying face, a state in which the packet loss has occurred and deteriorating of data is significant is indicated.

The invention discloses a three-dimensional exaggerated face generation method based on a single irony portrait; the method comprises the following steps: obtaining a deformation expression set of a normal face data set through expression face data and template face deformation on the normal face data set; combining a computer interactive mode with a normal face characteristic point detection model so as to extract a characteristic point coordinate set of an input irony portrait; according to a preset corresponding relation between the characteristic point coordinate set and a three-dimensional exaggerated face characteristic point coordinate set, combining a three-dimension to two-dimension projection process, and using characteristic points and deformation expression common constraints to optimize a three-dimensional exaggerated face complying with the image characteristic point information and having face properties. The method can prevent a tedious three-dimensional exaggerated face data set creating process; assisted by a new face expression, the method can effectively obtain a three-dimensional exaggerated face model only through a few characteristic point corrections.

The embodiment of the invention discloses an expression transformation method and device, electronic equipment and a computer readable medium. A specific embodiment of the method comprises the steps of obtaining a target face image; and inputting the target face image into a preset expression transformation model to obtain an expression transformation image. The expression transformation model carries out expression transformation on the target face image, thereby achieving different expression transformation effects. For example, smiling face special effect transformation (normal face imagesare transformed into corresponding smiling face special effect images) can be carried out. Besides, the locally processed face image set displaying the preset expression is used for training, so thatthe effect of additional special effects can be achieved on the basis of expression transformation image transformation.

The invention provides a method for detecting a distance between a human face and a camera based on image acquisition. The method includes acquiring a normal face image of a primitive human face positioned at a preset first axial distance and a preset second axial distance through a monocular camera, acquiring a first average pixel value and a second average pixel value corresponding to an human eye pupil distance in the normal face image, calculating an original mapping relationship between the human eye pupil distance and the human face distance according to the preset first axial distance,the preset second axial distance, the first average pixel value and the second average pixel value, acquiring the human face image to be detected by the monocular camera, and carrying out image processing on the human face image to obtain the eye pupil distance to be detected; and calculating the distance of the human face to be detected according to the original mapping relationship and the humaneye pupil distance to be detected. According to the method for detecting the distance between the human face and the camera based on image acquisition, the original mapping relationship between the human eye pupil distance and the human face distance is obtained, the distance between the human face to be detected is obtained according to the eye pupil distance to be detected and the original mapping relationship, and a distance measuring sensor is not needed in the whole process.

The invention is the vortex tube nozzle and belongs to the refrigeration field. The characteristics are: the flow channel (1) of the nozzle adopts the geometrical symmetry, which is that the central axis line of the flow channel is designed by the equal Mach grads and makes the air flow increase along the axis line to reduce the lost of the air flow; the air intake front channel (2) of the nozzle adopts the design of the same plane with the nozzle channel (1) and the equal air speed, which is to ensure the speed of the normal face of the central line of the air channel consistent to the intake of the vortex tube. The invention improves the refrigeration temperature effect of the vortex tube and increases the chill flow ratio of the maximum refrigeration to improve the unit refrigeration volume and the refrigeration coefficient.

The invention discloses a mask-worn face recognition method, which comprises the following steps of: acquiring a real-time picture by an equipment camera, inputting the picture into a trained mask detection model, and if mask wearing is detected, inputting a portrait into an eye detection model, and inputting a detected eye image into an eye feature extraction network to obtain eye features; and if the matched mask is not detected, inputting the portrait into the face feature extraction network to obtain face features, comparing cosine similarity between the features and a corresponding feature library, and outputting a recognition result meeting conditions. According to the method, the difference between the face wearing the mask and the normal face is considered, the recognizable area is fully utilized, and the feature library is separated, so that the equipment can recognize the normal face and the face wearing the mask, and the recognition precision is improved.

The invention discloses an autism spectrum disorder face screening method based on machine learning, and the method specifically comprises the following steps: obtaining a plurality of groups of facesample images which comprise a normal face image and an ASD face image; preprocessing the plurality of groups of face sample images; and inputting the preprocessed face sample image into a pre-constructed convolutional neural network model for training and testing to obtain an ASD face screening model. According to the method, the ASD face features are extracted by using the convolutional neural network, so that rapid recognition of the autism spectrum disorder image is realized, and compared with manual recognition, the image recognition accuracy is greatly improved.

The invention discloses a three-classification face detection method using context information. The method comprises the followings steps of 1) data preparation, 2) model design in which an end-to-endneural network model including a basic convolutional network, a regional proposal network and a fine adjustment network is designed; and 3) model training in which the neural network model designed in the step 2) is input via a training set, the model is trained in a batch random gradient decrease manner, a verification set is used to verify a training effect of the model, and finally a face detection model is obtained. The step 1) further comprises that a) each face in an image set is marked manually to obtain a face frame; b) the face frames after marking are classified according to the size, a small face area is amplified, and others are marked as normal faces; and c) a face image set with marks obtained in the step b) is divided into the training set and the verification set. Thus, the face detection precision and the face recall rate are improved greatly.

The invention discloses a flanging die. The flanging die comprises an upper die base, a lower die base, a flanging male die mounted at the bottom of the upper die base and flanging female dies mounted on the lower die base. The inner side faces of the flanging female dies are molding surfaces. The top end of each molding surface is provided with a molding rod. The molding rods are transversely inlaid in the molding surfaces correspondingly and can rotate on the corresponding molding surfaces. The opposite sides of the two molding rods are provided with rolling surfaces protruding out of the respective molding surfaces. The rolling surfaces are matched with the flanging male die to conduct flanging on blanks. According to the scheme, the rolling surfaces are used for replacing flanging female dies and are matched with the flanging male die to conduct flanging on parts. Linear contact molding (common surface contact) is avoided by adopting the rolling surfaces in the flanging process, so that flanged surfaces are more smooth, and the flanging molding effect is good.

The invention discloses a forged face change enhancement detection method based on head posture deviation correction. The method comprises the following steps: acquiring a normal face image and a forged face image and extracting key points; for the forged face image, combining with the head three-dimensional model, carrying out re-projection on the key points to obtain re-projected key points, correcting the projected key points, and according to the extracted initial key points and the corrected key points, carrying out distortion on the forged face image based on a least moving square algorithm to obtain an enhanced forged face image; taking the normal face image and the enhanced counterfeit face image as samples, performing head posture evaluation based on key points of the samples to determine a rotation matrix and a translation matrix of the head posture, and taking the rotation matrix and the translation matrix as construction features to optimize counterfeit face change detection model parameters; and performing enhanced detection of fake face change by using the parameter-optimized fake face change detection model. According to the method, the detection capability of a forged face change detection model is enhanced by enhancing a forged sample.

The invention relates to the field of artificial intelligence, in particular to a facial paralysis degree evaluation system based on artificial intelligence. The system comprises an image acquisitionmodule used for acquiring a to-be-detected image; an image cutting module which is used for cutting the to-be-detected image to obtain a to-be-detected face image; a to-be-tested face 3D model construction module which is used for constructing a to-be-tested face 3D model; a normal face 3D model matching module which is used for screening a normal face 3D model similar to the to-be-detected face in shape; an expression adjusting module which is used for adjusting expression parameters of the normal face 3D model to obtain a standard face 3D model; an ROI region division module which is used for dividing ROI regions for the to-be-detected human face 3D model and the standard human face 3D model, numbering the ROI regions and distributing corresponding weights, and calculating the distance between the central points of the two ROI regions with the same number; and a facial paralysis degree evaluation module which is used for evaluating the facial paralysis degree according to the weightand the distance. According to the system, the problem that errors exist when the facial paralysis degree is judged by utilizing the left-right symmetry of the human face in the prior art due to incomplete symmetry of the normal human face is solved.

The invention provides a method for synthesizing a face backlight picture. The method comprises the following steps: S1, preprocessing an image; s2, face feature point detection: S2.1, a normal face image is sent to a face detection model, and face contour points are detected; s2.2, inputting the background image into a face detection model, and detecting background face contour points; s3, generating a closed face contour line; s4, mask construction: S4.1, applying a flooding filling algorithm to the image template M to generate a face region mask 0; s4.2, according to the corresponding contour points, solving homography matrixes of the two images; homography transformation is carried out on the mask0; a mask 1 after transformation is obtained; homography transformation is carried out on the face image; a transformed face image is obtained; s4.3, carrying out negation operation on the mask 1, and generating a background region mask 2; s5, target area extraction: S5.1, performing mask operation on the transformed face image and a mask 1, and extracting a face area to obtain an image I1; s5.2, performing mask operation on the background image and a mask 2, and extracting a backlight background area to obtain an image I2; and S6, image splicing: carrying out addition operation on the obtained image pixels to obtain a spliced image.

The embodiment of the invention provides a three-dimensional exaggerated cartoon face generation method and device. The method comprises the following steps: inputting a normal face picture into a 3D-CariGAN model, outputting a PCA coefficient of a three-dimensional exaggerated cartoon face, and recovering the geometric model of the three-dimensional exaggerated cartoon face according to the PCA coefficient based on the PCA model of the three-dimensional exaggerated cartoon face; reconstructing a three-dimensional normal face according to the normal face picture, projecting the three-dimensional normal face to the normal face picture, and obtaining texture mapping of the three-dimensional exaggerated cartoon face; inputting the normal face picture into the CariStyGAN, and outputting the texture map of the three-dimensional exaggerated cartoon face; and obtaining the three-dimensional exaggerated cartoon face according to the geometric model, the texture mapping and the texture mapping.According to the embodiment of the invention, the natural complete three-dimensional exaggerated cartoon face is automatically generated by utilizing the PCA model of the three-dimensional exaggerated cartoon face.

The invention belongs to the technical field of access control systems, and particularly relates to an intelligent access control epidemic prevention system based on face recognition. The system comprises an initialization interface, a user management module, a mask detection module and a face recognition module. Each functional module is displayed in an interface after the initialization interface is entered, and the modules are clicked to realize functions; and the user management module has background administrator authority to manage resident information, and can perform information entry,information query, information modification and information deletion functions on the user in the background. According to the intelligent access control epidemic prevention system based on face recognition, the face recognition technology is used for recognizing a mask and a normal face, then judgment output is carried out at the decision level, the whole system can achieve the intelligent access control system with multi-layer safety control, and therefore the safety of traditional access control is improved, and the intelligent access control epidemic prevention system based on face recognition has high application value.

The embodiment of the invention discloses an abnormal face recognition method, an abnormal face recognition device, abnormal face recognition equipment and a storage medium, and the method comprises the steps: obtaining abnormal face data; processing the abnormal face data to obtain training face data, and extracting a first face feature in the training face data; obtaining recognized face data, and obtaining a second face feature in the recognized face data; embedding the second face feature into the first face feature, and constructing a recognition face data set; and training the recognizedface data set to realize abnormal face recognition. According to the embodiment of the invention, data processing is carried out through abnormal face data (abnormal face data), the face features inthe abnormal face data are enhanced, the feature data (recognized face data) of the normal face are embedded into the abnormal face feature data set through the LLE algorithm, deep training and modeloutput of the network are completed, and accurate abnormal face recognition can be achieved.

The invention relates to the technical field of behavior analysis, in particular to a student learning behavior analysis system based on big data, which comprises a server and a monitoring group arranged in a classroom, and is characterized in that the server stores score data of each student in each period, a score difference threshold, normal face information of each student and corresponding identity information; the server is used for identifying the identity of the student according to the picture uploaded by the monitoring group, a face analysis module is used for identifying students ina sleeping state and a distracting state through the facial features of the students, and marking the students with marks of the sleeping students and the distracting students; a post analysis moduleis used for identifying students with poor sitting postures through the postures of the students, and marking the students as students with ill-conditioned sitting postures; a comparison module is used for periodically comparing scores of students and calculating score difference values; and when the score difference value is greater than the score difference threshold, a difference analysis module is used for calculating the difference between each piece of marked data and the previous period, so as to solve the problem that the teacher is limited in energy in class and then the specific reason for rising and falling of the student learning scores is difficult to find, and the method is mainly used for schools.

The invention discloses a living body face detection model training method, which is applied to the technical field of artificial intelligence and is used for solving the technical problem that the prediction precision of a model does not reach the standard when a binary classification model is trained through extremely unbalanced positive and negative samples. The method provided by the invention comprises the following steps: acquiring a face picture sample set carrying a label; inputting a normal face image sample and an abnormal face image sample into a living body face detection model to be trained to obtain an original prediction value; correcting the original predicted value of the abnormal face picture sample to obtain a corrected predicted value; taking the corrected predicted value as a predicted value of the abnormal face picture sample, taking an original predicted value as a predicted value of the normal face picture sample, and training the living body face detection model to be trained according to the category to which the corresponding face picture sample identified in the label actually belongs. When a loss function of the living body face detection model converges, the trained living body face detection model is obtained.

The present invention provides a micromechanical treatment system for facial paralysis, the system comprising: a signal collector embedded in the healthy side for collecting movement signals of the healthy side; a micromachine embedded or partially embedded in the affected side; The terminal processor is used to receive and process the data collected by the signal collector and at the same time issue instructions to the micromachine; the system is aimed at unilateral facial paralysis, and the healthy side facial movement is normal, and the terminal processor monitors the normal facial movement of the healthy side, Send commands to the micromachine to pull the affected side to move, thereby ensuring the normal movement of the affected side.

The invention provides a face mask, and relates to an article for daily life. The face mask comprises a facial towel, wherein hanging racks are arranged at the upper ends of two sides of the facial towel; a convex support is arranged on a position in correspondence to a nose bridge; and ropes are arranged at the lower end of the face mask. The hanging racks, which are similar to spectacle frames, can be hung on ears; by virtue of the convex support, compression on the nose bridge can be relieved; and by virtue of the ropes, the lower end of the facial towel is fixed, so that the face mask, in comparison with a normal face mask, is more comfortable to wear.

The invention relates to a face image recognition method and device, a computer device and a storage medium. The method comprises the steps of using the computer device to firstly obtain a to-be-identified image; adopting a preset image recognition model to carry out face feature recognition on the to-be-recognized image, wherein the image recognition model is a face recognition model obtained bytraining an image with a Yin-Yang face and a normal face image in advance. Therefore, according to the face image recognition method provided by the invention, the recognition accuracy is relatively higher.

The invention is suitable for the field of computer vision, and provides a mask robust face recognition network and method, electronic equipment and a storage medium, the face recognition network comprises a whole image feature extraction network, a feature segmentation module, a first branch network and a second branch network, the whole image feature extraction network is used for extracting shallow whole image features from an input face image, the feature segmentation module is used for performing spatial segmentation on the shallow whole image features according to positions of preset segmentation points to obtain upper and lower shallow features, the first branch network is used for extracting upper half advanced features of upper half shallow features, the upper half advanced features are used for mask face recognition, the second branch network is used for extracting the lower half advanced features of the lower half superficial layer features, the lower half advanced features are used for being spliced with the upper half advanced features, and the spliced full features are used for normal face recognition, so that the robustness of the face recognition network to mask faces is improved.

The invention discloses an intelligent prompt method based on face recognition and a related device, so that intelligent equipment is more humanized, and the user experience is improved. The method comprises the following steps: acquiring a face image of a user through a camera shooting function of intelligent equipment, wherein the face image is a visible light picture containing the face of the user; recognizing the current face state of the user according to the face image; comparing the current face state with a normal face state of the user; if the current face state is not consistent with the normal face state, determining the current emotional state of the user through face feature information data in a pre-created face feature database; and generating corresponding intelligent prompts according to the current emotional state, wherein the intelligent prompts are different intelligent prompts corresponding to different preset face changes.

The embodiment of the invention discloses an image processing method, device and apparatus, and a storage medium. The method can comprise the steps that: a color face image and a depth face image comprising a target user face are obtained, and the color face image and the depth face image have an association relation, wherein the association relation means that the color face image and the depth face image are used for reflecting the same face form of a target user in the same scene; feature extraction is performed on the color face image to obtain a first feature; denoising processing is carried out on the depth face image, and normal information calculation is carried out on the depth face image after denoising processing to obtain a normal face image; feature extraction is performed onthe normal face image to obtain a second feature; feature fusion processing is performed on the first feature and the second feature to obtain a fused feature; and target identity information corresponding to the target user is determined based on the fused feature. By adopting the embodiment of the invention, the accuracy of face recognition can be improved.

The invention relates to an electric hot water bottle and a preparation method used for the electric hot water bottle. The electric hot water bottle of the invention comprises an independent software internal sleeve and an external sleeve; liquid is memorized in the internal sleeve; the edge of the internal sleeve is also provided with a first connecting part; the position on the external sleeve corresponding to the first connecting part is provided with a second connecting part which is matched with the first connecting part; the first connecting part and the second connecting part are fixedly connected directly with each other between the internal sleeve and the external sleeve by a fixed part, thus leading the connection of the edge of the internal sleeve and the corresponding external sleeve to be non-splittable structure. The electric hot water bag of the structure adopts an independent software internal sleeve, namely, the internal sleeve is individually formed; furthermore, the edge of the internal sleeve and the external sleeve are fixedly connected with each other by the fixed part in a non-splittable way; therefore, the external sleeve can select normal face fabric cloth as the material and has wide selection range; the preparation cost is reduced, furthermore, the type and pattern of the normal face fabric cloth can be designed as diversity so as to enhance the sense of beauty.

The invention relates to the technical field of image privacy protection, and provides a method for protecting the privacy of a human face image and electronic equipment, and the method comprises the steps: carrying out the formatted representation of an original human face image, obtaining an original image matrix of pixel points of at least one human face position needing privacy processing, and carrying out the privacy processing through employing a trained human face generation model, the difference noise matrix can keep the distribution of the visual generation matrix of the key part consistent with the distribution of the original image matrix, so that the privacy of the original image matrix is perturbed, and the privacy of the original image matrix is perturbed. After the face local image generated by the vision generation matrix of the key part is used for replacing the corresponding key part in the original face image, it can be ensured that the obtained target face image is a normal face image, that is, good vision is kept; moreover, the target face image is tested, so that the usability of the target face image is ensured, and the personal privacy of the user is also protected.

The invention discloses a facial feature migration method, electronic equipment and a storage medium. The method comprises the following steps: acquiring an updated first feature extraction model; if the to-be-recognized image does not exist at present, performing feature extraction on the image in the image library by using a first feature extraction model to obtain a first face feature, and storing the first face feature in a first feature library; in the process of performing feature extraction by the first feature extraction model, if a new to-be-recognized image exists, performing face recognition on the new to-be-recognized image by using features in the first feature library and / or the second feature library, wherein the second feature library is obtained by performing feature extraction on the images in the image library by using the second feature extraction model before updating. By means of the mode, face feature migration can be achieved, and normal face recognition is not affected.