195 results about "Face identity" patented technology

The invention provides a 3D (three-dimensional) face identity authentication method and device. The method includes the steps: acquiring a depth image and a two-dimensional image comprising a target face; registering the depth image with a reference face 3D texture image to obtain pose information of the target face; aligning the two-dimensional image according to the pose information to obtain atarget face two-dimensional image; extracting characteristic information in the target face two-dimensional image; comparing similarities of the characteristic information in the target face two-dimensional image and characteristic information in a reference face two-dimensional image. The pose of the target face is acquired according to the 3D information, alignment is performed according to thepose, consistency of the current target face two-dimensional image and the reference face two-dimensional image is ensured to a greater degree, and identification accuracy is improved. Besides, the method further includes the steps of human eye sight line detection, detection in vivo and data updating to improve user experience, reduce false acceptance rate and deal with the problems of face change and the like.

The invention discloses an infrared face identification method based on a local parallel nerve network. A network structure mainly comprises four portions and the method is characterized by 1, extracting a preliminary convolution characteristic: through one group of 2*2 convolution kernel, extracting a preliminary face characteristic and arranging an output characteristic signal; 2, generating a parallel multi-scale convolution characteristic: using a parallel multi-scale convolution network structure to extract face characteristics representing different scale information; 3, generating a classification characteristic vector: using a fully-connected layer to integrate the convolution characteristic so as to acquire a characteristic vector which is used to finally represent a face identity and is used for classification input, and carrying out modified linear activation and random neglect processing; and 4, training and testing a classifier: inputting a processed fully-connected characteristic vector into a Softmax classifier and calculating losses, and reversely spreading, training and adjusting a network parameter so as to realize infrared face identification. The method can be widely applied to infrared face identification and identity identification applications.

The invention provides a 3D human face quick identity authentication method and apparatus. The method comprises the following steps of S1, obtaining a depth image and a two-dimensional image comprising a current user face; S2, obtaining the depth image and the two-dimensional image of the current user face; S3, performing identity authentication primary screening on the current user face to obtaincandidate reference data sets; S4, extracting feature information of the current user face, and determining a pose orientation vector of the current user face; and S5, performing identity authentication. The functions of input, detection, identification and the like of a face identity are realized by utilizing the depth image and the two-dimensional image, and the primary screening of the reference data sets is performed in combination with the depth image and the two-dimensional image, so that the identification efficiency is improved, the influence of an external complex environment is avoided, and absolute front face of a user is not required.

The invention discloses an identity labeling method of face images and a face identity recognition method of the face images. The face identity recognition method includes the steps of (1) labeling the identities of the face images to be labeled: searching the face images similar to the face images and corresponding webpages, determining the identities of the face images according to the frequencies of appeared names in the returned webpages, detecting the identities of the face images respectively through a face technology platform and a face identity recognition model, and synthesizing the recognition results to determine the final identities of the face images and label the face images, (2) carrying out matching filtering on a set of face images belonging to the same names and the face images with the label results as the names in the step (1), (3) extracting feature vectors of the filtered identity labeled face images, training the labeled face images with a machine learning algorithm, and generating a face identity recognition model, and (4) as for two face images to the detected, extracting the feature vectors of the face images to judge whether the two face images belong to the same person or not through the face identity recognition model. According to the identity labeling method and the face identity recognition method, the labeling efficiency and the recognition effect are greatly improved.

The invention discloses a face abstracting method based on face recognition. The face abstracting method comprises the steps of generating face images of different persons appearing in an original video and forming an arisen face image list. The method further comprises the steps of face detection, face feature extraction, face feature clustering, face abstract image generation and the like, wherein the face detection is used for scanning image frames in the original video and judging whether face areas exist in the video frames. The invention further discloses a device for implementing the face abstracting method. According to the face abstracting method based on face recognition and the device thereof, the face images of different persons appearing in the video can be generated, the generated face image of each person is the most representative image appearing in the video and the corresponding position information in the original video. The face images can be rapidly recognized manually, and can also be used as input data of a current face identity recognition system or a face identity recognition system with the higher precision in the future.

The invention is applicable to the field of face recognition technology and provides a face recognition method and device based on deep learning. The method includes the steps of constructing a deep neural network based on face training images; acquiring a to-be-recognized image; detecting a face region in the to-be-recognized image and then extracting the face region; converting the face region image into a standard frontal face image and inputting the converted image into the deep neural network; outputting an expression vector of the standard frontal face image by using the deep neural network; and comparing the expression vector with each face description feature in a face library to obtain the face identity of the to-be-recognized image. In the invention, since the plurality of face training images are used as supervisory information to establish the deep neural network and character features of each image are extracted based on the deep neural network, the character features withhigher robustness can be learned and used, and compared with a traditional face recognition method, the method in the invention has better face recognition effect and can have a stronger anti-interference capability under complicated environment conditions.

The invention discloses a method for constructing a face recognition model based on a ParaSoftMax loss function, which comprises the following steps: selecting a basic convolution neural network modelaccording to an application environment of a task; acquiring a face image marked with human face identity information in a specified number as a training data set; the decision edge parameters are obtained according to the difference of the class center angles between the difficult sample eigenvectors and the simple sample eigenvectors and the class center angles in the basic convolution neural network model. Obtaining a ParaSoftMax loss function according to the decision edge parameter; setting the loss function at the last layer of the basic convolution neural network model to form a face recognition model based on the loss function; input the training data set to the face recognition model, minimizing the loss function iterative training model parameters, and obtaining the optimal facerecognition model. Thus, the face recognition model of the present application can improve the accuracy of face recognition.

The present invention discloses a system and method of detection in vivo. The method comprises: detecting each face area in many face images of users and extracting the face key points corresponding to the face motions to be detected from the detected face areas; determining whether the face image are continuous or not; and performing detection in vivo of the face image based on the extracted face key points to determine that the face images come from real people, where in the step of execution of detection in vivo of face images includes the motion detection step, the texture detection step and the three-dimensional mode detection step. The present invention further discloses a face identity identification system and method with silence detection in vivo. The face identity identification system and method with silence detection in vivo integrate a plurality of modes of detection in vivo, automatically identify the users' motions or expressions and perform special detection in vivo of the motions or expressions so as to improve the accuracy of the detection in vivo.

The invention provides a face verification anti-counterfeit recognition method and a system thereof based on an interactive action. The method comprises a step of carrying out the initial recording of the information of an register static face image and the information multiple register face action images, a step of waiting the reading of a static face image to be read, matching a character and a stored character when the shooting of static face image to be detected obtained by a user to be verified is detected, and if a matching degree reaches the storage characteristic of a preset threshold value, conforming to a verification requirement, and a step of randomly selecting and prompting the user to be verified to complete a corresponding face action according to the recorded face action, extracting the characteristic of the action image to be detected of the user be verified, matching the characteristic with the historical verification feature information of a corresponding face action, completing face identity verification if a matching rate reaches a preset threshold value, adding the matched image into the historical verification feature information, returning to select a next face action to continue matching if the matching is not approved or does not reaches a desired effect, treating the verification as a failure if the action exceeds a preset number of times, and ending the verification.

The invention provides a method for building a deep learning based face recognition and age synthesis joint model. The method is characterized in that alignment and PCA and LDA dimension reduction preprocessing are performed on a pair of input images; six groups of features for identity representation and different age group representation are acquired through an automatic encoder acquired by training, an image similarity degree is outputted for each of the six results through a parallel CNN, then a matching result is acquired through weighted fusion. The method provided by the invention has an excellent effect for independent face recognition or age detection or a common task, and can also acquire an excellent effect for face recognition under the influence of illumination and postures; and the method also has robustness for cross-age face recognition because features of the age and the face identity are separated. In addition, some parameters and weights can be adjusted according to requirements, so that the method is very flexible.

The invention provides a 3D face identity authentication method and device. The method comprises the following steps: obtaining a depth image and a two-dimensional image of a target face; matching thedepth image with a reference face 3D texture image to obtain posture information of the target face; projecting a reference face two-dimensional image via the reference face 3D texture image according to the posture information; comparing similarity between the target face two-dimensional image and the reference face two-dimensional image. Through combination of 3D information and projection, themethod is capable of obtaining complete reference face two-dimensional image so as to improve identification accuracy; meanwhile, the method also comprises the steps of detecting human eye sight, detecting living body and updating data so as to improve user experience, reduce false identification rate and deal with the problems such as the face changes.

The invention discloses a face attribute analysis-based person and certificate integrated identity authentication method. According to the method, person and certificate consistency authentication is carried out based on race, gender, age and facial features presented on an on-site face image. The method includes the following steps of: on-site face image acquisition; certificate information acquisition; and person and certificate integrated authentication, wherein the on-site face image acquisition step further includes face detection, face image normalization and face image quality evaluation, the certificate information acquisition step further includes race, gender, age and face facial information extraction, and the person and certificate integrated authentication step further includes race authentication, gender authentication, age authentication and face comparison authentication. According to the method, prior information such as race, gender and age on the certificate is fully utilized, and therefore, obvious fraudulent use of the certificate under in an unattended scene can be rapidly filtered out; and criminals can be prevented from passing identity authentication by cheating a camera through using a printed image at the surface of the certificate.

The invention provides a 3D face identity authentication method and an apparatus. The method comprises the following steps of acquiring a depth image sequence and a two-dimensional image sequence containing a target face; calculating a 3D texture image of the target face; using the 3D texture image to project a two-dimensional image of the target face; extracting characteristic information in thetwo-dimensional image of the target face; and carrying out similarity comparison on the characteristic information in the two-dimensional image of the target face and characteristic information in a two-dimensional image of a reference face. In the method, 3D information and projection are combined to acquire a complete two-dimensional image of the target face so as to increase identification precision; simultaneously the method comprises steps of human eye sight detection, in vivo detection and data updating so that a user experience is improved, a false identification rate is reduced, and face changes and other problems can be solved.

The invention discloses a face recognition system, which relates to the technical field of image processing and pattern recognition. First, the gender of a face image to be recognized is determined through a deep learning method. Then, the face identity of the face image to be recognized is determined with high accuracy according to the gender identified through deep learning and based on a convolutional neural network. Finally, the facial micro expression of the face image to be recognized is determined through a binocular vision technology based on the parallax characteristic of the face image to be recognized. Face recognition based on the gender information, identity information and facial expression information of a face image to be recognized is realized. By combining the deep learning method, the convolutional neural network and the binocular vision technology, the influence of factors such as illumination and facial gesture on the face recognition rate is reduced, the amount of computation in the face recognition process is reduced, fast and high-accuracy face recognition is realized, the cost of face recognition is reduced, and the accuracy of face recognition under different facial postures is improved.

The invention discloses an ATM terminal human face key points partially shielding detection method based on a random forest comprises the following steps: acquiring an image when a bank card is inserted into an ATM, improving the gray image contrast degree through a multi-scale Retinex enhancing method, positioning a human face through adoption of a Haarcascade human face detection algorithm, speculating a human face gesture based on a gesture estimation tree model of the random forest, judging a human face shielding type through a Viola Jones frame, and, according to the estimated human face gesture, selecting a random forest key point detection model and calibrating a human face key point of a position which is not shielded. An influence of the gesture on human face key point detection is reduced through gesture pre-estimation based on the random forest, an influence of shielding on a human face characteristic is reduced through automatic recognition of a shielding portion and key point marking of a non-shielding portion, a high-precision human face local characteristic can be provided for ATM-based identity recognition of a shielded human face, and human face identity recognition precision is improved.

The invention discloses an attitude robustness face recognition method based on deep learning. The problem that in the prior art, the face recognition accuracy and speed with attitude changes need to be improved is solved. The method comprises the following steps: step 1, preprocessing a training sample; step 2, constructing and training a face identity recognition network; step 3, constructing and training a head posture recognition network; and step 4, constructing and training a feature fusion network. Firstly, face identity characteristics and posture characteristic information are extracted by using a convolutional neural network, characteristic fusion is carried out on the two kinds of information, and finally, cosine similarity measurement is carried out on fusion characteristics containing the face identity information and the posture information, and whether the fusion characteristics belong to the same person or not is judged, thereby finishing face identity recognition. The invention discloses a face recognition technology with posture robustness through a feature fusion method, and the face recognition accuracy and speed with posture change are improved.

On a synthesis task of a human face, a multilevel sparse expression three-time conversion virtual generation neural network TTGAN is constructed based on an adversarial generation network CycleGAN architecture. The TTGAN proposes and joins a multi-level sparse representation model and a three-time conversion consistency constraint, and the TTGAN is a result under the synergistic effect of a plurality of generative adversarial networks for the target face synthesis of a face image pair. Wherein the multi-level sparse representation model is used for constraining features extracted by differentfeature extraction layers of a generated network in an input picture, including identity information related to a target image; The three times of conversion consistency constraint utilizes three different samples which contain network state information and are generated by one time of circulation of the model, so that the two generative adversarial networks of the whole model are guided to cooperate with each other. The multi-level sparse representation and the three-time conversion consistency constraint provided by the TTGAN further increase the image generation capability of the CycleGAN,so that the synthesized face image can obtain a better result in the aspects of keeping face identity information and showing more reality.

The invention discloses a face recognition method and a face recognition device based on fusion of multiple frames of face features in a video. The face recognition method comprises the following steps: acquiring n frames of face images to be recognized in a monitoring video, wherein n is greater than or equal to 1; selecting m frames of face images from the n frames of face images, and performingfeature extraction on the m frames of face images to generate feature vectors {fi} in one-to-one correspondence with the m frames of face images, wherein i is 1, 2, ...., m, and m is greater than orequal to 1 and smaller than or equal to n; fusing the m feature vectors {fi} into a feature vector r, and comparing the feature vector r with face features in a database so as to recognize a face identity in the monitoring video. By the face recognition method provided by the invention, multiple frames of face images in the monitoring video are detected, feature extraction is performed thereon, and the extracted multiple face features are fused into one face feature for recognition, so that not only is the number of feature comparisons reduced, but also influence of face angle deflection, motion blur, backlight and the like on face image feature extraction is reduced; the face recognition method and the face recognition device are applied to a multi-dynamic video acquisition environment and can effectively improve the face recognition accuracy.

The invention discloses a face replacement method based on a multi-level attribute encoder and an attention mechanism. The face replacement method mainly solves the problems that in the prior art, target attributes such as background and illumination are ignored for image replacement, and the fusion effect is poor. According to the scheme, the method comprises the following steps: 1) preprocessing a source face image by using a multi-task convolutional neural network; 2) extracting source face identity features through a feature encoder; 3) extracting target face image attributes by using a multi-level attribute encoder through multi-level cascaded convolution blocks and deconvolution blocks and interlayer connection; 4) constructing a novel generator network in combination with an attention mechanism, and designing a generator loss function; 5) making a network training set and a test set, and performing iterative training on the novel generator network; and 6) generating a face replacement image by using the trained network model. The method can comprehensively and accurately extract the attributes of the target image, better preserves the posture, expression, illumination and other information of the target face, and generates a real and natural face replacement image.

The invention discloses a vehicle-mounted intelligent identity recognizing and monitoring system, and relates to the field of electronic monitoring technology. The system comprises an automobile starting control system and a face recognition system, wherein the face recognition system comprises a camera, a microprocessor module and a memory module; the camera is connected with the microprocessor module; the memory module is connected with the microprocessor module; the microprocessor module analyzes and compares face characteristic data of a driver, which is captured by the camera with the face characteristic data of permitted drivers, which is stored in the memory module; the microprocessor module is connected with the automobile starting control system; and the accordance between the face characteristic data of the driver and the face characteristic data of the permitted driver, which is stored in the memory module is used as one of the control parameters for automobile starting. The characteristic of the driver is recognized through face identity recognition and in the vehicle-mounted intelligent identity recognizing and monitoring system, the face characteristic of the driver is taken into consideration, and the face characteristic of a person is obvious, so the anti-theft capability is strong.

The invention discloses a face recognition algorithm training method based on multi-objective learning, comprising the following steps: randomly initializing neural network parameters, using a loss function based on face identity and a loss function based on face feature point position constraint to minimize the learning objective to train the deep convolution network; when the prediction accuracyrate of the face identity reaches the threshold, calculating the loss function based on the distance within the face feature classes and the loss function based on the distance between the face feature classes, and performing calculation of the loss function based on the face identity and the loss function of the face feature point position constraint on each sample; based on the artificial weight setting, weighting each loss function to obtain the total loss function, and realizing the back propagation based on the total loss function. Therefore, the update of the network parameters is realized, and the network training is stopped when the accuracy is stable to get the trained face recognition model.

The invention relates to the field of image recognition, in particular to a face recognition method and device, computer equipment and a storage medium. Identifying the to-be-identified face image byusing a deep neural network model to obtain various kinds of face feature information in the to-be-identified face image; And outputting the various kinds of face feature information. According to theinvention, image features are extracted through the convolutional neural network; the high self-learning ability is realized; The method has the advantages that the robustness is high, the lightweight network is used for extracting image features, the generalization is better, the recognition speed is higher, the multi-task learning thought is adopted, the bottom convolutional neural network shares parameters, and the upper network exclusively shares specific parameters, so that the recognition task of the face identity, gender and age can be completed by one model, the model is simplified, and the efficiency is high; Information mined in face identity recognition can be continuously used in gender recognition and age recognition, and gender and age recognition precision is improved.

The invention discloses a human face recognition method taking a convolutional neural network (CNN) as a feature extractor. The method is characterized by comprising the following steps of SS1: constructing a to-be-identified human face image database according to a recognition need, and forming a training set and a test set according to needs; SS2: extracting a CNN model as the feature extractor; SS3: extracting to-be-identified eigenvectors of images of the to-be-identified human face library by the CNN model, and performing storage; and SS4: classifying extracted training set and test set data by utilizing a feature classifier. According to the method, a relatively popular deep learning CNN in existing image recognition serves as the feature extractor, so that the characteristics of high recognition rate and high anti-jamming property of the CNN can be utilized, the to-be-identified human face database can be conveniently subjected to modification operation, and the practicality of the human face recognition method in the field of human face identity recognition is improved.

The invention provides a face identification method applied to an adaptive drive seat; the method comprises the following steps: loading a face identity characteristic head portrait database; building a face identification model, and training a loaded face identification database; obtaining a video image from a camera; using a cascade classifier to detect whether the video image contains face characteristic information or not; if yes, extracting the face portion, and forming a face image; carrying out dimension normalization for the extracted characteristic head portrait, and carrying out histogram equalization treatment; using two dimension discrete rapid Fourier transform to convert the face image from a space domain to a frequency domain, and extracting characteristics; comparing the extracted characteristics in the face identification database, if the similarity is higher than a preset threshold, a predicted ID label is outputted so as to confirm the passenger ID, thus starting the adaptive drive seat; asking to input the face identity if the similarity is smaller than the preset threshold. The face identification method is applied to an unmanned vehicle auxiliary driving system, and matched with the adaptive drive seat, thus fast and accurately carrying out face identification with high efficiency.

The invention belongs to the technical field of pattern recognition, and in particular relates to a composite gradient vector-based face recognition method. The method comprises the following steps of: marking a target area in a positioned face image, dividing feature subareas in the target area, performing orthogonal sampling by using marginal singular points of the feature subareas as starting points and end points of vectors to obtain base vectors, constructing all the base vectors in the target area into a vector cluster, performing multi-dimensional compounding on the base vectors to obtain all great gradient vectors in the vector cluster, constructing a composite gradient vector by using the great gradient vectors as elements, counting the dimension and the gradient information of the composite gradient vector, and comparing the composite gradient vector and the dimension and the gradient information of the composite gradient vector with a face library to recognize face identity. Compared with other face recognition methods, the face recognition method provided by the invention has the advantages of stronger environmental suitability and feature extraction capacity and high recognition performance under the conditions of illumination intensity variation, multiple gestures and multiple expressions and can be used for face recognition under the large-range complex environment in the field of biological feature identification.

The invention relates to a user identity recognition and matching method based on face recognition technology. Through user device clients, faces of users are automatically acquired to serve as face identity characteristic images which are recorded in a system server. Target user identity images shot by user devices are allowed to match the face identity characteristic images recorded in the system server and user identities corresponding to the target user identity images are determined. According to the invention, by searching the face identity characteristic images which are recorded in thesystem server and match all users in position regions where the users are located, the searching range is shortened, search speed is improved and accuracy of recognition and matching is improved.