21 results about "Anti spoofing" patented technology

The invention discloses an intelligent face recognition system based on a swan-gap access control system. Comprising a multi-modal biological characteristic acquisition module, a self-adaptive illumination compensation module, an edge-cloud cooperative computing module, a behavior intention analysis module, a privacy compliance processing module, an anti-AI cheating protection module, a self-adaptive authority management module and an energy consumption optimization and self-maintenance module. And data communication and task scheduling among the modules are realized through a swan gap distributed soft bus, and the system has the beneficial effects that the security, efficiency and privacy protection capability of the system are remarkably improved through the multi-mode biological characteristic acquisition module, the dynamic illumination compensation module, the edge-cloud cooperative computing module and the like. The anti-AI cheating protection and the living body detection guarantee the authenticity of recognition, and the fake face attack is avoided; the energy consumption optimization and self-maintenance module prolongs the service life of hardware and reduces the long-term operation and maintenance cost; the self-adaptive authority management and emergency escape priority mechanism enhances the emergency response capability.

A method for detection of a spoof call comprises: receiving a first voice response in response to a received call from a caller's device; augmenting the first voice response with a first watermark signal; sending the augmented first voice response to the caller's device; receiving a second voice response from the caller's device, wherein the second voice response is expected to contain therein an echo of the first watermark signal; analyzing the second voice response to determine the existence of the first watermark signal in the second voice response; and, generating a spoof call indication upon determination of the absence of the first watermark in the second voice response. System, circuit and non-transitory computer-readable medium embodiments are further provided.

Some disclosed methods involve providing a user prompt to place a user's digit in a fingerprint sensor area and obtaining, from a fingerprint sensor, fingerprint sensor data corresponding to the user's digit. Some disclosed methods involve determining, based on the fingerprint sensor data, that a currently-obtained fingerprint image matches one or more previously-obtained fingerprint images. Some disclosed methods involve extracting, from the fingerprint sensor data, N top-ranked anti-spoofing fingerprint feature types of a set of T anti-spoofing fingerprint feature types, N and T being integers and N being an integer smaller than T. Some disclosed methods involve performing an anti-spoofing process using the N top-ranked anti-spoofing fingerprint feature types.

An anti-spoofing authentication system includes an image capture device that converts received light into an image signal representing a captured image; a microphone that converts sound wave into a voice signal; a face detection device that detects a human face on the captured image, the face detection device including a lip detection device that detects lip features associated with the detected human face; a voice activity detection (VAD) device that detects presence of human voice in the voice signal; and an anti-spoofing device that detects a spoofing attack according to the detected lip features and the detected human voice.

The invention belongs to the technical field of artificial intelligence and computer vision, and particularly relates to a face anti-cheating detection method and system based on enhanced adapter fine tuning, and the method comprises the following steps: collecting a face image data set, carrying out the data preprocessing, and dividing the obtained data set into a source domain and a target domain; respectively training and verifying the ViT target detection network model of the integrated enhanced anti-spoofing adapter EAS-Adapter by using the source domain and the target domain to obtain a face anti-spoofing detection model; wherein the ViT target detection network model is that an EAS-Adapter is integrated on each Transform coding layer of a Transform coder in the ViT network, and the EAS-Adapter is integrated on each Transform coding layer of the Transform coder in the ViT network; and performing data preprocessing on a to-be-detected face image, and inputting the to-be-detected face image into the face anti-deception detection model to obtain a face authenticity identification result. According to the invention, each Transform encoding layer of the Transform encoder is integrated with the EAS-Adapter to capture complementary deception clues, and high-precision face authenticity identification in a cross-domain scene is realized.

Various aspects presented herein may enable a UE to identify whether an image captured by a camera of the UE is spoofing (e.g., is a real image or a false / manipulated / virtual image, etc.). In one aspect, a UE obtains a set of images associated with a vision assisted positioning session, where the set of images is captured using at least one first camera. The UE detects the presence of at least one spoofing feature in the set of images during the visually assisted positioning session. The UE stores or outputs an indication of the at least one spoofing feature based on the presence of the at least one spoofing feature in the set of images. The UE performs the camera-assisted positioning session based on at least one non-spoofing feature, wherein the at least one non-spoofing feature is different from the at least one spoofing feature.

The invention discloses a face anti-deception domain generalization method based on style hybrid recombination, and the method is characterized in that the method comprises the following steps: S1, constructing a data processing model; s11, constructing an adversarial learning module; s12, constructing a comparative learning module; s13, constructing a style assembly layer; s14, constructing a loss function; and S2, a training process. The invention relates to the technical field of image processing, in particular to a face anti-deception domain generalization method based on style hybrid recombination. The technical problem to be solved by the invention is to provide a face anti-deception domain generalization method based on style hybrid recombination, which is characterized in that a complete feature is separated into a content feature and a style feature, the content feature and the style feature are respectively processed, the content feature is indistinguishable among different domains through adversarial learning, and the content feature and the style feature are integrated. Style features emphasize information related to activity through comparative learning, and meanwhile, domain-specific information is inhibited. According to the method, the generalization ability of the model in the unseen domain is effectively improved.

The invention belongs to the technical field of network security, and discloses a user identification method and device, equipment, a storage medium and a product. According to the method, deception person prediction is carried out on all users through a preset deception person prediction identification model, deception users are obtained, then, the comprehensive scores of the camouflage persons corresponding to the camouflage persons are determined according to call information corresponding to the deception users, and the final scores of the camouflage persons are determined according to the comprehensive scores of the camouflage persons and the scores of an anti-deception center. And determining a risk population corresponding to the camouflage-prone person according to the final score of the camouflage-prone person. According to the method and the device, the deception user in the total amount of users can be obtained, then the comprehensive score of the camouflage person is determined according to the call information corresponding to the deception user, and the final score of the camouflage person is determined according to the comprehensive score of the camouflage person and the score of the anti-deception center, so that compared with an existing method of screening the camouflage person by an expert, the method and the device have the advantages that the workload is reduced; according to the method, the risk crowd corresponding to the camouflage-susceptible person can be determined according to the final score of the camouflage-susceptible person, so that the risk crowd susceptible to cheating can be automatically and accurately identified.

Aspects presented herein may enable a UE to identify whether images captured by the camera(s) of the UE are spoofed (e.g., are real images or false / manipulated / virtual images, etc.). In one aspect, a UE obtains a set of images associated with a vision-aided positioning session, where the set of images is captured using at least one first camera. The UE detects that at least one spoofing feature is present in the set of images during the vision-aided positioning session. The UE stores or outputs an indication of the at least one spoofing feature based on the at least one spoofing feature being present in the set of images. The UE performs the camera-aided positioning session based on at least one non-spoofing feature, where the at least one non-spoofing feature is different from the at least one spoofing feature.

Aspects presented herein may enable a UE to verify the integrity of map data, thereby improving the accuracy and security of map-assisted positioning and / or map-based positioning. In an aspect, the UE performs map-assisted positioning based on a first set of map data. The UE verifies whether the integrity of the first set of map data satisfies an accuracy threshold. The UE discards the first set of map data if the verification of the integrity of the first set of map data does not meet the accuracy threshold. For example, the UE may compare a first set of map data to a second set of map data from a different source, and identify that the integrity of the first set of map data does not meet an accuracy threshold if the comparison shows an indication of an inconsistency above a consistency threshold.

The application provides a navigation anti-deception method, device and electronic equipment, wherein the method comprises the following steps: constructing a network topology structure according to a communication topology constraint and a space geometric relation of a formation network, wherein the formation network establishes time-frequency synchronization and unified coordinate conversion; selecting a plurality of nodes from the network topology structure to construct an anti-deception array; calculating distances between array elements in the anti-deception array and a signal source; comparing the calculated distances between the array elements and the signal source with actual height information of satellites to determine whether the received signal is an interference signal; constructing a positioning equation based on the anti-deception array and measurement parameters thereof, and solving the positioning equation by using a least square method to obtain position coordinates of the interference signal. The application can adapt to deception interference detection and interference source positioning under different terminal layouts and task requirements, and has lower cost and is efficient and reliable.

A Driver Monitoring System, DMS, configured to detect spoofing attempts, the system comprising: an NIR image sensor module; a light source; a controller configured to control the light source and the image sensor module, so to capture each consecutive image of the sequence in a different lighting condition; a processor having processor operable software, the software comprising: a motion analyzer, operable to analyze a motion in the sequence of images, and output a motion-based result; a first neural network, operable to extract at least one type of information from an at least two channel image created by stacking a set of at least two consecutive images from the image sequence, and output a spatial result; and a fusion program, operable to merge the motion-based and the spatial results, and to output a final result, the final result determining if the target is live or fake.

The present invention discloses a GNSS anti-spoofing detection method based on a damped accumulation and particle swarm optimization gray model, comprising the following steps: obtaining vehicle-mounted IMU data and odometer data between any two frames of GPS signals, using the previous frame GPS data as a benchmark, using the above data to infer and record the vehicle's geographic coordinate position when each frame of IMU data arrives, that is, the longitude, latitude and altitude time series, as input to a gray model optimization algorithm based on damped accumulation and particle swarm optimization; using the optimization algorithm to predict the vehicle's geographic coordinate position when the next frame GPS signal arrives, as the algorithm output; after coordinate system conversion, using the Euler formula to calculate the distance between the predicted geographic coordinates and the geographic coordinates obtained by GPS; using the distance obtained in the previous step to compare with the range threshold to determine whether the vehicle has been deceived. The present invention improves the robustness of the vehicle's GPS and further enhances the vehicle's anti-spoofing capability for GPS deception signals by optimizing the prediction algorithm.

The invention belongs to the technical field of authority verification, and particularly relates to an AI big data elevator control access control management system and method, by fusing multi-channel sensing data, collaborative analysis of multi-modal features such as human faces, postures and gaits is realized, a spatio-temporal behavior mode and environment context cross validation mechanism is introduced, and physical layer anti-cheating detection is combined, so that the security of the elevator control access control management system is improved. And the authenticity of the access request is comprehensively evaluated. According to the scheme, multi-source perception, dynamic behavior modeling and a physical layer anti-deception mechanism are fused, so that high recognition accuracy can be maintained under the severe conditions of low illumination, partial shielding and the like, advanced deception means such as image playback, confrontation samples, 3D masks and the like can be effectively resisted, abnormal access opportunities and trailing behaviors can be recognized, and the recognition accuracy is improved. Therefore, the overall safety level and the intelligent level of the access control system are remarkably improved, and the technical defect that a traditional system is prone to being bypassed or misjudged in a complex application scene is overcome.

The application claims a face anti-spoofing method based on feature consistency and scale pyramid loss, belonging to the technical field of computer vision. The method comprises the following steps: first, in order to solve the problem that the model generalization is reduced due to the insufficient feature consistency under different viewing angles in living body detection, a feature consistency module is proposed to encourage the consistency of feature maps under different viewing angles and ensure that feature maps under different scales can be matched and aligned with each other. Secondly, a module MFFM for multi-scale feature fusion is designed, so that the field migration part can learn more scale information when generating a depth map, which helps the network to better capture the correlation and spatial structure between features. Finally, a scale pyramid loss SP Loss is designed, which can make the model consider the information of the image from multiple scales, including global semantics and local details, to improve the perception ability of the model to different scale information and further improve the performance and robustness of the model.

Disclosed herein is a field of biometric authentication, in particular the field of anti-spoofing. Further disclosed herein are methods, apparatuses, devices, material information and computer elements for generating biometric authentication information.

A network nodes pairing method having an anti-spoofing mechanism is provided used in a mesh network system. Each of node apparatuses in the mesh network system uses group information of the mesh network system as a key to encrypt network system identification information and self-identification information to broadcast a beacon signal including encrypted information. A scan process is executed by a child node apparatus to select a target father node. A validity verification process is performed on the beacon signal of the target father node according to the group information so as to determine the target father node passes the validity verification process when the target father node and the child node apparatus has the same network system identification information and when the self-identification information of the target father node is valid. A pairing request signal is transmitted to the target father node to request for pairing.

The present invention discloses a domain generalization method for face anti-spoofing based on style hybrid recombination, characterized in that it includes the following steps: S1: constructing a data processing model; S11: constructing an adversarial learning module; S12: constructing a contrastive learning module; S13: constructing a style assembly layer; S14: constructing a loss function; S2: training process. The present invention relates to the field of image processing technology, and in particular, to a domain generalization method for face anti-spoofing based on style hybrid recombination. The technical problem to be solved by the present invention is to provide a domain generalization method for face anti-spoofing based on style hybrid recombination, and proposes to separate complete features into content features and style features, and process them separately. The content features are made indistinguishable between different domains through adversarial learning, while the style features are emphasized with respect to activity-related information through contrastive learning, while suppressing domain-specific information. This method effectively improves the generalization ability of the model in unseen domains.

Techniques and apparatuses are described that implement pattern recognition for face-authentication anti-spoofing. In particular, a face-authentication system distinguishes between a real human face and a presentation attack that uses a display to present a virtual human face. Operational settings of a camera system of the face-authentication system are tailored to enable detection of a pattern associated with an operation of the display. A spoofing detector of the face-authentication system analyzes one or more images captured by the camera system and determines whether or not the pattern is present within the image(s). If the pattern is present, the face-authentication system does not provide face authentication. Alternatively, if the pattern is not present, the face-authentication system can provide face authentication if facial recognition is successful. In this way, the face-authentication system can prevent unauthorized actors from using the presentation attack to gain access to a user's account or information.

The invention discloses a voice anti-spoofing detection method and device. The method comprises the steps that high-dimensional features corresponding to an audio to be detected are acquired, and the high-dimensional features comprise multi-scale identification information of the audio to be detected; a nested architecture is adopted to process the high-dimensional features to obtain target features, the nested architecture comprises multi-level nested residual connection and a multi-scale feature fusion mechanism, and the target features comprise enhanced feature representation obtained by interaction of multi-scale identification information; and performing anti-deception detection on the target feature to obtain a detection result corresponding to the to-be-detected audio, the detection result including the probability that the to-be-detected audio is a deception audio. According to the invention, the technical problem that the detection performance is limited due to the loss of discriminative information in original features caused by the introduction of a dimension reduction layer in a voice anti-cheating detection method adopted in the related technology is solved.

A method includes obtaining dual-pixel image data that represents an object and includes a first sub-image and a second sub-image, and generating (i) a first feature map based on the first sub-image and (ii) a second feature map based on the second sub-image. The method also includes generating a correlation volume by determining, for each respective offset of a plurality of offsets between the first feature map and the second feature map, pixel-wise similarities between (i) the first feature map and (ii) the second feature map offset from the first feature map by the respective offset. The method further includes determining, by an anti-spoofing model and based on the correlation volume, a spoofing value indicative of a likelihood that the object represented by the dual-pixel image data is being spoofed.