CN114821732A

CN114821732A - Living body detection method, model training method, device, electronic device, and medium

Info

Publication number: CN114821732A
Application number: CN202210511476.6A
Authority: CN
Inventors: 黄泽斌
Original assignee: Beijing Baidu Netcom Science and Technology Co Ltd
Current assignee: Beijing Baidu Netcom Science and Technology Co Ltd
Priority date: 2022-05-10
Filing date: 2022-05-10
Publication date: 2022-07-29

Abstract

The disclosure provides a living body detection method, a model training method, a living body detection device, an electronic device and a medium, relates to the technical field of artificial intelligence, specifically to the technical field of deep learning, image processing and computer vision, and can be applied to scenes such as face recognition. The specific implementation scheme is as follows: acquiring a face region image and an extended region image of the target object, wherein the face region image includes an image of a first face region, and the extended region image is an image of a second face region and a predetermined extended region other than the second face region; obtaining target image data according to the face area image and the at least one extended area image of the target object, wherein the target image data comprises at least one of the following: a first fused image and auxiliary image data, the auxiliary image data comprising at least one of: a contour image and a second fused image; and performing living body detection by using the target image data to obtain a living body detection result aiming at the target object.

Description

Living body detection method, model training method, device, electronic device, and medium

Technical Field

The disclosure relates to the technical field of artificial intelligence, in particular to the technical field of deep learning, image processing and computer vision, and can be applied to scenes such as face recognition. In particular, it relates to a method for detecting a living body, a method for training a model, an apparatus, an electronic device, and a medium.

Background

With the continuous development of face recognition technology, in various identity verification scenes, the face recognition technology is increasingly popularized to authenticate and apply user identities. For the scene of identity authentication by adopting the face recognition technology, in addition to face recognition, living body detection is required. Live body detection can be used to determine whether the acquired face image is from a live body object.

Disclosure of Invention

The disclosure provides a living body detection method, a model training method, a living body detection device, an electronic device and a medium.

According to an aspect of the present disclosure, there is provided a method of living body detection, including: acquiring a face region image and an extended region image of a target object, wherein the face region image includes an image of a first face region, and the extended region image is an image of a second face region and a predetermined extended region other than the second face region; obtaining target image data according to the face region image and the extended region image, wherein the target image data includes at least one of: a first fused image and auxiliary image data, said auxiliary image data comprising at least one of: a contour image and a second fused image; and performing a biopsy using the target image data to obtain a biopsy result for the target object.

According to another aspect of the present disclosure, there is provided a training method of a living body detection model, including: acquiring a sample face area image of a sample object and a sample extended area image, wherein the sample face area image is an image of a first sample face area, and the sample extended area image is an image of a second sample face area and a predetermined sample extended area other than the second sample face area; obtaining sample image data according to the sample face area image and the sample extended area image, wherein the sample image data includes at least one of: a first sample fused image and sample auxiliary image data, said sample auxiliary image data comprising at least one of: the sample outline image and the second sample fusion image; and training a deep learning model by using the sample image data to obtain the living body detection model.

According to another aspect of the present disclosure, there is provided a living body detection apparatus including: a first acquisition module configured to acquire a face region image of a target object and an extended region image, wherein the face region image is an image of a first face region, and the extended region image is an image of a second face region and a predetermined extended region other than the second face region; a first obtaining module, configured to obtain target image data according to the face region image and the extended region image, where the target image data includes at least one of: a first fused image and auxiliary image data, said auxiliary image data comprising at least one of: a contour image and a second fused image; and a second obtaining module, configured to perform living body detection by using the target image data, so as to obtain a living body detection result for the target object.

According to another aspect of the present disclosure, there is provided a training apparatus of a living body detection model, including: a second acquisition module configured to acquire a sample face area image of a sample object and a sample extended area image, wherein the sample face area image is an image of a first sample face area, and the sample extended area image is an image of a second sample face area and a predetermined extended area other than the second sample face area; a fifth obtaining module, configured to obtain sample image data according to a sample face area image of a sample object and at least one sample extended area image, where the sample image data includes at least one of: a first sample fused image and sample auxiliary image data, said sample auxiliary image data comprising at least one of: the sample outline image and the second sample fusion image; and the sixth obtaining module is used for training a deep learning model by utilizing the sample image data to obtain the living body detection model.

According to another aspect of the present disclosure, there is provided an electronic device including: at least one processor; and a memory communicatively coupled to the at least one processor; the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor to enable the at least one processor to perform the method of the present disclosure.

According to another aspect of the present disclosure, there is provided a non-transitory computer readable storage medium having stored thereon computer instructions for causing the computer to perform the method of the present disclosure.

According to another aspect of the present disclosure, a computer program product is provided, comprising a computer program which, when executed by a processor, implements the method of the present disclosure.

It should be understood that the statements in this section do not necessarily identify key or critical features of the embodiments of the present disclosure, nor do they limit the scope of the present disclosure. Other features of the present disclosure will become apparent from the following description.

Drawings

The drawings are included to provide a better understanding of the present solution and are not to be construed as limiting the present disclosure. Wherein:

FIG. 1 schematically illustrates an exemplary system architecture to which the liveness detection method, the training method of the liveness detection model, and the apparatus may be applied, according to an embodiment of the present disclosure;

FIG. 2 schematically illustrates a flow chart of a liveness detection method according to an embodiment of the present disclosure;

FIG. 3 schematically illustrates a flow chart for acquiring a face region image and an extended region image of a target object according to an embodiment of the present disclosure;

FIG. 4 schematically illustrates a flow chart for deriving target image data from a face region image and an extended region image according to an embodiment of the present disclosure;

FIG. 5 schematically illustrates a flowchart for in vivo detection using target image data resulting in vivo detection results for a target object, in accordance with an embodiment of the present disclosure;

FIG. 6 schematically illustrates an example schematic diagram of a liveness detection process according to an embodiment of the disclosure;

FIG. 7 schematically illustrates a flow chart of a method of training a liveness detection model according to an embodiment of the present disclosure;

FIG. 8 schematically illustrates a block diagram of a liveness detection device according to an embodiment of the present disclosure;

FIG. 9 schematically illustrates a block diagram of a training apparatus for a liveness detection model according to an embodiment of the present disclosure; and

fig. 10 schematically shows a block diagram of an electronic device adapted to implement a liveness detection method and a training method of a liveness detection model according to an embodiment of the present disclosure.

Detailed Description

Exemplary embodiments of the present disclosure are described below with reference to the accompanying drawings, in which various details of the embodiments of the disclosure are included to assist understanding, and which are to be considered as merely exemplary. Accordingly, those of ordinary skill in the art will recognize that various changes and modifications of the embodiments described herein can be made without departing from the scope and spirit of the present disclosure. Also, descriptions of well-known functions and constructions are omitted in the following description for clarity and conciseness.

The living body detection can verify whether the object is a living body object by combining actions of blinking, mouth opening, shaking head, nodding head and the like and by utilizing technologies of human face key point positioning, human face tracking and the like. The living body detection can effectively resist the attacking means such as photos, face changing, masks, sheltering, screen copying and the like, thereby ensuring the benefits of the object.

The embodiment of the disclosure provides a living body detection scheme. For example, a face region image of the target object including an image of a first face region and an extended region image that is an image of a second face region and a predetermined extended region other than the second face region are acquired; obtaining target image data according to the face area image and the extended area image, wherein the target image data comprises at least one of the following data: a first fused image and auxiliary image data, the auxiliary image data comprising at least one of: a contour image and a second fused image. And performing living body detection by using the target image data to obtain a living body detection result aiming at the target object.

Fig. 1 schematically illustrates an exemplary system architecture to which the in-vivo detection method, the training method of the in-vivo detection model, and the apparatus may be applied, according to an embodiment of the present disclosure.

It should be noted that fig. 1 is only an example of a system architecture to which the embodiments of the present disclosure may be applied to help those skilled in the art understand the technical content of the present disclosure, and does not mean that the embodiments of the present disclosure may not be applied to other devices, systems, environments or scenarios. For example, in another embodiment, an exemplary system architecture to which the living body detection method and the training method and apparatus for a living body detection model can be applied may include a terminal device, but the terminal device may implement the living body detection method and the training method and apparatus for a living body detection model provided by the embodiments of the present disclosure without interacting with a server.

As shown in fig. 1, the system architecture 100 according to this embodiment may include

terminal devices

101, 102, 103, a network 104 and a server 105. The network 104 serves as a medium for providing communication links between the

terminal devices

101, 102, 103 and the server 105. Network 104 may include various connection types, such as wired and/or wireless communication links, and so forth.

The user may use the

terminal devices

101, 102, 103 to interact with the server 105 via the network 104 to receive or send messages or the like. The

terminal devices

101, 102, 103 may have installed thereon various communication client applications, such as a knowledge reading application, a web browser application, a search application, an instant messaging tool, a mailbox client, and/or social platform software, etc. (by way of example only).

The

terminal devices

101, 102, 103 may be various electronic devices having a display screen and supporting web browsing, including but not limited to smart phones, tablet computers, laptop portable computers, desktop computers, and the like.

The server 105 may be various types of servers that provide various services. For example, the Server 105 may be a cloud Server, which is also called a cloud computing Server or a cloud host, and is a host product in a cloud computing service system, so as to solve the defects of high management difficulty and weak service extensibility in a conventional physical host and VPS service (Virtual Private Server). Server 105 may also be a server of a distributed system or a server that incorporates a blockchain.

It should be noted that the living body detection method provided by the embodiment of the present disclosure may be generally executed by the

terminal device

101, 102, or 103. Accordingly, the living body detection apparatus provided by the embodiment of the present disclosure may also be provided in the

terminal device

101, 102, or 103.

Alternatively, the living body detection method provided by the embodiments of the present disclosure may also be generally performed by the server 105. Accordingly, the living body detecting apparatus provided by the embodiments of the present disclosure may be generally disposed in the server 105. The liveness detection method provided by the embodiments of the present disclosure may also be performed by a server or server cluster that is different from the server 105 and that is capable of communicating with the

terminal devices

101, 102, 103 and/or the server 105. Accordingly, the living body detection apparatus provided by the embodiment of the present disclosure may also be provided in a server or a server cluster different from the server 105 and capable of communicating with the

terminal devices

101, 102, 103 and/or the server 105.

It should be noted that the training method of the living body detection model provided by the embodiment of the present disclosure may be generally executed by the server 105. Accordingly, the training device of the living body detection model provided by the embodiment of the present disclosure may be generally disposed in the server 105. The training method of the model of the living body detection model provided by the embodiment of the present disclosure may also be performed by a server or a server cluster that is different from the server 105 and is capable of communicating with the

terminal devices

101, 102, 103 and/or the server 105. Accordingly, the training apparatus of the living body detection model provided by the embodiment of the present disclosure may also be disposed in a server or a server cluster different from the server 105 and capable of communicating with the

terminal devices

101, 102, 103 and/or the server 105.

Alternatively, the training method of the living body detection model provided by the embodiment of the present disclosure may also be generally executed by the

terminal device

101, 102, or 103. Accordingly, the training apparatus of the living body detection model provided by the embodiment of the present disclosure may also be disposed in the

terminal device

101, 102, or 103.

It should be understood that the number of terminal devices, networks, and servers in fig. 1 is merely illustrative. There may be any number of terminal devices, networks, and servers, as desired for implementation.

It should be noted that the sequence numbers of the respective operations in the following methods are merely used as representations of the operations for description, and should not be construed as representing the execution order of the respective operations. The method need not be performed in the exact order shown, unless explicitly stated.

FIG. 2 schematically shows a flow chart of a method of in-vivo detection according to an embodiment of the disclosure.

As shown in FIG. 2, the method 200 includes operations S210-S230.

In operation S210, a face region image of a target object including an image of a first face region and an extended region image that is an image of a second face region and a predetermined extended region other than the second face region are acquired;

in operation S220, target image data is obtained according to the face region image and the extended region image, the target image data including at least one of: a first fused image and auxiliary image data, the auxiliary image data comprising at least one of: a contour image and a second fused image.

In operation S230, a live body detection is performed using the target image data, resulting in a live body detection result for the target object.

According to an embodiment of the present disclosure, the target object may refer to an object that needs to be subjected to living body detection. The object may include at least one of: a living subject and a non-living subject. A living object may refer to an object having a life. For example, the living subject may include a human body. A non-living object may refer to an object that has no life. The non-living object may include an attack object. The attack object may be a "live object" forged by an attack method. The attack pattern may include at least one of: print attacks, display attacks, and mask attacks. A print attack may refer to printing a target object on an image. Thus, the target object in the print image can be referred to as an attack object. A display attack may refer to displaying a video including a target object on a display screen. Thus, a target object in the video may be referred to as an attack object. A mask attack may refer to a mask obtained from a target object. Thus, the target object in the mask image can be referred to as an attack object. Based on the above, the image may include one of: live images and non-live images. The living body image may refer to an image corresponding to a living body object. The non-living object image may be an image corresponding to a non-living object. The non-live image may include an attack image. The attack image may include at least one of: print images, video frame images, and mask images.

According to an embodiment of the present disclosure, the target object may include a face region and other regions other than the face region with respect to the target object. The other region may include a predetermined extension region. The predetermined area may include the second face area and a predetermined extension area outside the second face area. The image of the target object corresponding to the first face region may be referred to as a face region image. An image of the target object corresponding to the predetermined area is referred to as an extended area image. The first and second face regions may be the same or different. The predetermined extension area may be configured according to actual service requirements, and is not limited herein. For example, the predetermined expansion region may be a peripheral region of the face region. The predetermined extension area may include at least one of: the hair region, ear region, neck region, upper limb region, predetermined region may be a predetermined multiple of the face region. The predetermined multiple may be configured according to actual service requirements, and is not limited herein. For example, the predetermined multiple may be one of: 2 times, 3 times and 4 times.

According to an embodiment of the present disclosure, at least one extended area image may be included. The respective extended area images are different from each other. The target image data may be a result of fusing the face region image and the at least one extended region image. The face region image and the extended region image may be derived from a scene image. The living body detection result for the target object may include one of: the target object is a living object and the target object is a non-living object.

According to the embodiment of the present disclosure, a scene image of a target object may be acquired. The scene image may be processed to obtain a face region image and at least one extended region image of the target object. The face region image and the at least one extended region image may be processed to obtain target image data. For example, the face region image and the at least one extended region image may be fused to obtain a first fused image. The fusing may include one of splicing and adding. Processing may be performed based on the first fused image to obtain auxiliary image data. The auxiliary image data may be used to assist the first fused image to determine a living body detection result of the target object. Accuracy of a result of the living body detection of the target object is improved based on the auxiliary image data. Target image data is obtained based on at least one of the first fused image and the auxiliary image data. For example, the contour image may be a contour of the first fused image. The second fused image may be color space converted from the first fused image. According to the embodiments of the present disclosure, after the target image data is obtained, the living body detection may be performed on the target image data, resulting in a living body detection result for the target object. For example, the target image data may be processed using a living body detection method based on a deep learning model, resulting in a living body detection result for the target object. The deep learning model may be one of: supervised, semi-supervised and unsupervised models. The deep learning model may include one of: a ResNet (Residual Neural Network) model, a VGG (Visual Geometry Group) model, a MobileNet model, a LeNet model, and an AlexNet model. Alternatively, the target image data may be processed using a live body detection method based on manual features, resulting in a live body detection result for the target object. For example, the living body detection method based on the texture information may process the target image data to obtain a living body detection result. Alternatively, the living body detection method based on the image quality may process the target image data to obtain the living body detection result. Alternatively, the target image data may be processed based on a three-dimensional structure live body detection method to obtain a live body detection result.

According to an embodiment of the present disclosure, the above operations S210 to S230 may be performed by an electronic device. The electronic device may comprise a terminal device or a server. For example, the terminal device may be terminal device 101, terminal device 102, or terminal device 103 in fig. 1. The server may be the server 105 in fig. 1.

According to the embodiments of the present disclosure, the target image data includes the face region image of the target object and the extended region image, which is an image including the second face region of the target object and a predetermined extended region image other than the second face region, and therefore, the target image data may include information of a non-face region in addition to information of the face region, thereby making the information included in the target image data more comprehensive, and on the basis of this, the accuracy of the living body detection result can be improved by performing the living body detection using the target image data.

The in-vivo detection method according to the embodiment of the disclosure is further described with reference to fig. 3 to 6.

Fig. 3 schematically shows a flowchart of acquiring a face region image and at least one extended region image of a target object according to an embodiment of the present disclosure.

As shown in fig. 3, the method 300 is further defined as operation S210 in fig. 2, the method 300 including operations S311-S312.

In operation S311, an image of a first face region is extracted from a scene image of a target object, resulting in a face region image.

In operation S312, an image of the second face region and a predetermined extended region other than the second face region is extracted from the scene image, resulting in an extended region image.

According to an embodiment of the present disclosure, the scene image may be an image in one of the following formats: RGB images, BGR images, YUV images, HSV images, and infrared images. R can be characterized (Red ). G characterization (Green ). B characterization (Blue ). Y can be characterized (Luminance). U can be characterized (chroma). V in YUV can be characterized (Chroma). H can be characterized (Hue). S can be characterized (Saturation). V in HSV can be characterized (Value ). The format of the scene image may be configured according to actual service requirements, and is not limited herein.

According to embodiments of the present disclosure, the scene image may be acquired in real time. For example, the scene image may be acquired using a vision sensor. Alternatively, the scene image may be obtained by searching the image set for an image related to the target object using a web crawler method. The obtaining mode of the scene image may be configured according to actual business requirements, and is not limited herein.

According to an embodiment of the present disclosure, a scene image including a target object may be acquired in response to an image acquisition request. Location information of a face region in a scene image may be determined. A first face region image is extracted from the scene image based on the position information of the face region. For example, the position information may include coordinate information of each of four vertices of the position frame. The four vertices may include an upper left vertex, a lower left vertex, an upper right vertex, and a lower right vertex. The location frame may comprise a rectangular frame.

According to the embodiment of the disclosure, for at least one predetermined area, the position information of the predetermined area can be determined, and the image of the predetermined area is extracted from the scene image according to the position information of the predetermined area, so as to obtain the extended area image. The predetermined area may include the second face area and a predetermined extension area outside the second face area. For example, the position information of the predetermined region in the scene image may be determined based on the position information of the face region in the scene image and a predetermined magnification ratio. The coordinate information may be multiplied by a predetermined magnification ratio for each of the coordinate information of the four vertices of the face region in the scene image, to obtain magnified coordinate information. Four magnified coordinate information can thus be obtained. And extracting an extended area image from the scene graph according to the four pieces of enlarged coordinate information. The predetermined amplification ratio value may be a number greater than 1. The predetermined amplification ratio may be configured according to actual service requirements, and is not limited herein. For example, the predetermined amplification ratio may be one of: 2. 3 and 4. The at least one extended area image may include a scene image.

According to an embodiment of the present disclosure, the above operations S311 to S312 may be performed by an electronic device. The electronic device may comprise a terminal device or a server. For example, the terminal device may be terminal device 101, terminal device 102, or terminal device 103 in fig. 1. The server may be the server 105 in fig. 1.

Fig. 4 schematically shows a flowchart for obtaining target image data from a face region image and an extended region image according to an embodiment of the present disclosure.

As shown in fig. 4, the method 400 may be further defined as to operation S220 in fig. 2, the method 400 including operations S421 to S423.

In operation S421, the face region image and the extended region image of the target object are stitched to obtain a first fused image.

In operation S422, auxiliary image data is obtained based on at least one of the first fused image, the face region image, and the extended region image.

In operation S423, target image data is obtained according to at least one of the first fusion image and the auxiliary image data.

According to the embodiment of the disclosure, the face region image and the at least one extended region image may be stitched to obtain a first fused image. The auxiliary image data may be obtained, for example, from the first fused image. Alternatively, the auxiliary image data is derived from the face region image and the at least one extended region image. Alternatively, the auxiliary image data is derived from the first fused image, and the face region image and the at least one extended region image.

According to an embodiment of the present disclosure, the target image data may be obtained by, for example, determining the first fusion image as the target image data. Alternatively, the auxiliary image data is determined as target image data. Alternatively, the target image data may be derived from the first fused image and the auxiliary image data.

According to an embodiment of the present disclosure, the above operations S421 to S423 may be performed by an electronic device. The electronic device may comprise a terminal device or a server. For example, the terminal device may be terminal device 101, terminal device 102, or terminal device 103 in fig. 1. The server may be the server 105 in fig. 1.

Operation S422 in fig. 4 may include the following operations according to an embodiment of the present disclosure.

And processing the first fusion image by using a contour extraction algorithm to obtain a contour image. Obtaining a second fused image using one of: and carrying out color space conversion on the first fusion image to obtain a second fusion image. A second fused image is obtained from the converted face region image, which is obtained by color space converting the face region image, and at least one converted extended region image, which is obtained by color converting the extended region image.

According to an embodiment of the present disclosure, the contour extraction algorithm may include one of: the contour extraction algorithm based on the Sobel operator, the contour extraction algorithm based on the Scharr filter and the contour extraction algorithm based on the Canny operator. The Sobel operator can be used to determine the approximate gradient of the image gray scale function.

According to an embodiment of the present disclosure, the color space may include one of: RGB color space, BGR color space, YUV color space, and HSV color space. For example, the color space of the first fused image is an RGB color space. The first fused image may be converted from an RGB color space to an HSV color space, resulting in a second fused image. Alternatively, the face-area image may be converted from an RGB color space to an HSV color space, resulting in a converted face-area image. And converting the extended area image in the at least one extended area image from an RGB color space to an HSV color space to obtain at least one converted extended area image. And fusing the converted face region image and the at least one converted extended region image to obtain a second fused image.

Fig. 5 schematically shows a flowchart for performing a biopsy using target image data to obtain a biopsy result for a target object according to an embodiment of the present disclosure.

As shown in fig. 5, the method 500 may be further defined as operation S230 of fig. 2, the method 500 including operations S531-S532.

In operation S531, feature extraction is performed on the target image data to obtain target feature data.

In operation S532, living body classification is performed according to the target feature data, and a living body detection result for the target object is obtained.

According to the embodiment of the disclosure, the target image data can be processed by using the deep learning model, and the living body detection result for the target object can be obtained. The deep learning module may include a feature extraction module and a classification module. The classification module may include a classifier.

According to the embodiment of the disclosure, living body classification can be performed according to target feature data, and a classification probability value is obtained. And obtaining a living body detection result aiming at the target object according to the classification probability value and a preset probability threshold value. The predetermined probability threshold may be configured according to actual service requirements, and is not limited herein. For example, the predetermined probability threshold may be 0.8. In a case where it is determined that the classification probability value is greater than or equal to the predetermined probability threshold value, a living body detection result that the target object is a living body object is obtained. And obtaining a living body detection result of the target object which is a non-living body object under the condition that the classification probability value is determined to be smaller than a preset probability threshold value.

According to an embodiment of the present disclosure, the above operations S531 to S532 may be performed by an electronic device. The electronic device may comprise a terminal device or a server. For example, the terminal device may be terminal device 101, terminal device 102, or terminal device 103 in fig. 1. The server may be server 105 in fig. 1.

The in-vivo detection method according to the present disclosure is further described with reference to fig. 6 in conjunction with specific embodiments.

Fig. 6 schematically shows an example schematic diagram of a liveness detection process according to an embodiment of the present disclosure.

As shown in fig. 6, in 600, an image of a first face region is extracted from a scene image 601 of a target object, resulting in a face region image 602. For each of the two predetermined regions, an image of the predetermined region is extracted from the scene image, resulting in an extended region image, from which an extended region image 603 and an extended region image 604 can be obtained. The predetermined area may include the second face area and a predetermined extension area outside the second face area.

The face region image 602, the extended region image 603, and the extended region image 604 of the target object are stitched to obtain a first fused image 605.

The first fused image 605 is processed using a contour extraction algorithm to obtain a contour image 606.

From the first fused image 605 and the contour image 606, target image data 607 is obtained. Using the target image data 607, a living body detection result 608 for the target object is obtained.

Fig. 7 schematically shows a flowchart of a training method of a liveness detection model according to an embodiment of the present disclosure.

As shown in fig. 7, the method 700 includes operations S710 to S730.

In operation S710, a sample face area image of a sample object, which is an image of a first sample face area, and a sample extension area image, which is an image of a second sample face area and a predetermined sample extension area other than the second sample face area, are acquired.

In operation S720, sample image data is obtained from a sample face area image and at least one sample extended area image of a sample object, the sample image data including at least one of: a first sample fused image and sample auxiliary image data, the sample auxiliary image data comprising at least one of: the sample contour image and the second sample fused image.

In operation S730, a deep learning model is trained using the sample image data, resulting in a living body detection model.

According to the embodiments of the present disclosure, for the descriptions of the sample object, the sample face region image, and the sample extended region image, reference may be made to the above related contents for the target object, the face region image, and the extended region image, which are not described herein again.

According to the embodiment of the disclosure, the deep learning model may be configured according to actual business requirements, and is not limited herein. For example, the deep learning model may include one of: ResNet model, VGG model, MobileNet model, LeNet model and AlexNet model. The deep learning model may include at least one of: supervised, semi-supervised and unsupervised models.

According to the embodiment of the disclosure, the in-vivo detection model can be obtained by training the deep learning model by using the in-vivo detection result and the in-vivo detection label value of the sample object based on the loss function. The in-vivo detection model can be used for realizing in-vivo detection. The loss function may be configured according to actual service requirements, and is not limited herein. For example, the loss function may include at least one of: cross entropy loss function, exponential loss function, and squared loss function.

According to an embodiment of the present disclosure, the above operations S710 to S730 may be performed by an electronic device. The electronic device may comprise a server or a terminal device. For example, the server may be server 105 in FIG. 1. The terminal device may be terminal device 101, terminal device 102 or terminal device 103 in fig. 1.

According to the embodiment of the present disclosure, the sample image data includes the sample face area image of the sample object and the sample extended area image, the sample extended area image is an image including the second sample face area of the sample object and a predetermined sample extended area image other than the second sample face area, and therefore, the sample image data may include non-sample face area information in addition to information of the sample face area, thereby making information included in the sample image data more comprehensive, and on the basis of this, the deep learning model is trained using the sample image data, so that the living body detection can learn the features of the non-face area as well, thereby improving the accuracy of the living body detection result of the living body detection model. The sample auxiliary data is used for training the deep learning model, and the robustness of the in-vivo detection model can be improved.

Operation S720 may include the following operations according to an embodiment of the present disclosure.

And splicing the sample face area image and the sample expansion area image to obtain a first sample fusion image. Sample auxiliary image data is obtained from at least one of the first sample fusion image, the sample face area image, and the sample extension area image. Sample image data is obtained from at least one of the first sample fusion image and the sample auxiliary image data.

According to an embodiment of the present disclosure, obtaining the sample auxiliary image data from at least one of the first sample fusion image, the sample face area image, and the sample extension area image may include the following operations.

And processing the first sample fusion image by using a contour extraction algorithm to obtain a sample contour image.

Obtaining a second sample fusion image using one of:

and carrying out color space conversion on the first sample fusion image to obtain a second sample fusion image.

And obtaining a second sample fusion image according to the sample conversion face area image and the at least one sample conversion extension area image, wherein the sample conversion face area image is obtained by performing color space conversion on the sample face area image, and the sample conversion extension area image is obtained by performing color conversion on the sample extension area image.

Operation S730 may include the following operations according to an embodiment of the present disclosure.

And inputting the sample image data into the deep learning model to obtain the in-vivo detection result of the sample object. And inputting the in-vivo detection result of the sample object and the in-vivo detection label value into a preset loss function to obtain an output value. And adjusting the model parameters of the deep learning model according to the output value until the preset conditions are met to obtain the living body detection model.

According to an embodiment of the present disclosure, the predetermined condition may include at least one of convergence of the output value and a training round reaching a maximum training round.

According to an embodiment of the present disclosure, inputting sample image data into a deep learning model to obtain a biopsy result of a sample object may include: and carrying out feature extraction on the sample image data by using the deep learning model to obtain sample feature data. And carrying out living body classification on the sample characteristic data by using the deep learning model to obtain a living body detection result aiming at the sample object.

According to an embodiment of the present disclosure, the training method of the above-described living body detection model may further include the following operations.

An image of a first sample face area is extracted from a sample scene image of a sample object, and a sample face area image is obtained. And extracting images of the second sample face area and a preset expansion area except the second sample face area from the sample scene image to obtain a sample expansion area image.

In the technical scheme of the disclosure, the collection, storage, use, processing, transmission, provision, disclosure and other processing of the personal information of the related user are all in accordance with the regulations of related laws and regulations and do not violate the good customs of the public order.

The above is only an exemplary embodiment, but is not limited thereto, and other in-vivo detection methods and training methods of in-vivo detection models known in the art may also be included as long as the accuracy of the in-vivo detection result can be improved.

Fig. 8 schematically shows a block diagram of a living body detecting apparatus according to an embodiment of the present disclosure.

As shown in FIG. 8, the liveness detection device 800 may include a first acquisition module 810, a first acquisition module 820, and a second acquisition module 830.

A first obtaining module 810 for obtaining a face area image of the target object, the face area image being an image of the first face area, and an extended area image being an image of the second face area and a predetermined extended area other than the second face area.

A first obtaining module 820, configured to obtain target image data according to the face region image and the at least one extended region image of the target object, where the target image data includes at least one of: a first fused image and auxiliary image data, the auxiliary image data comprising at least one of: a contour image and a second fused image.

And a second obtaining module 830, configured to perform living body detection by using the target image data, so as to obtain a living body detection result for the target object.

According to an embodiment of the present disclosure, the first obtaining module 820 may include a first obtaining sub-module, a second obtaining sub-module, and a third obtaining sub-module.

And the first obtaining submodule is used for splicing the face region image and the extended region image to obtain a first fusion image.

A second obtaining sub-module for obtaining auxiliary image data based on at least one of the first fusion image, the face region image, and the extended region image.

And the third obtaining submodule is used for obtaining target image data according to at least one of the first fusion image and the auxiliary image data.

According to an embodiment of the present disclosure, the second obtaining sub-module may include a first obtaining unit, a second obtaining unit, and a third obtaining unit.

And the first obtaining unit is used for processing the first fusion image by utilizing a contour extraction algorithm to obtain a contour image.

Obtaining a second fused image using one of:

and the second obtaining unit is used for carrying out color space conversion on the first fusion image to obtain a second fusion image.

A third obtaining unit configured to obtain a second fusion image based on the converted face region image obtained by color space converting the face region image and the at least one converted extended region image obtained by color converting the extended region image.

According to an embodiment of the present disclosure, the second obtaining module 830 may include a fourth obtaining sub-module and a fifth obtaining sub-module.

And the fourth obtaining submodule is used for carrying out feature extraction on the target image data to obtain target feature data.

And the fifth obtaining submodule is used for carrying out living body classification according to the target characteristic data to obtain a living body detection result aiming at the target object.

According to an embodiment of the present disclosure, the above-mentioned living body detecting device 800 may further include a third obtaining module and a fourth obtaining module.

And the third obtaining module is used for extracting the image of the first face area from the scene image of the target object to obtain a face area image.

And the fourth obtaining module is used for extracting the second face area and an image of a preset expansion area except the second face area from the scene image to obtain an expansion area image.

Fig. 9 schematically shows a block diagram of a training apparatus of a living body detection model according to an embodiment of the present disclosure.

As shown in fig. 9, the training apparatus 900 for a living body detection model may include a second obtaining module 910, a fifth obtaining module 920, and a sixth obtaining module 930.

A second acquiring module 910, configured to acquire a sample face area image of the sample object and a sample extended area image, where the sample face area image is an image of the first sample face area, and the sample extended area image is an image of the second sample face area and a predetermined extended area other than the second sample face area.

A fifth obtaining module 910, configured to obtain sample image data according to the sample face area image and the sample extended area image, where the sample image data includes at least one of: a first sample fused image and sample auxiliary image data, the sample auxiliary image data comprising at least one of: the sample contour image and the second sample fused image.

A sixth obtaining module 930, configured to train the deep learning model with the sample image data to obtain the in-vivo detection model.

According to an embodiment of the present disclosure, the fifth obtaining module 920 may include a sixth obtaining sub-module, a seventh obtaining sub-module, and an eighth obtaining sub-module.

And the sixth obtaining submodule is used for splicing the sample face area image and the sample expansion area image to obtain a first sample fusion image.

A seventh obtaining sub-module for obtaining sample auxiliary image data from at least one of the first sample fusion image, the sample face area image, and the sample extension area image.

And an eighth obtaining submodule for obtaining sample image data from at least one of the first sample fusion image and the sample auxiliary image data.

According to an embodiment of the present disclosure, the sixth obtaining module 930 may include a ninth obtaining sub-module, a tenth obtaining sub-module, and an eleventh obtaining sub-module.

And the ninth obtaining submodule is used for inputting the sample image data into the deep learning model to obtain the in-vivo detection result of the sample object.

And the tenth obtaining submodule is used for inputting the living body detection result of the sample object and the living body detection label value into a preset loss function to obtain an output value.

And the eleventh obtaining submodule is used for adjusting the model parameters of the deep learning model according to the output value until the preset conditions are met, so that the living body detection model is obtained.

The present disclosure also provides an electronic device, a readable storage medium, and a computer program product according to embodiments of the present disclosure.

According to an embodiment of the present disclosure, an electronic device includes: at least one processor; and a memory communicatively coupled to the at least one processor; wherein the memory stores instructions executable by the at least one processor to enable the at least one processor to perform the method as described above.

According to an embodiment of the present disclosure, a non-transitory computer-readable storage medium having stored thereon computer instructions for causing a computer to perform the method as described above.

According to an embodiment of the disclosure, a computer program product comprising a computer program which, when executed by a processor, implements the method as described above.

Fig. 10 schematically shows a block diagram of an electronic device adapted to implement a liveness detection method and a training method of a liveness detection model according to an embodiment of the present disclosure. Electronic devices are intended to represent various forms of digital computers, such as laptops, desktops, workstations, personal digital assistants, servers, blade servers, mainframes, and other appropriate computers. The electronic device may also represent various forms of mobile devices, such as personal digital processing, cellular phones, smart phones, wearable devices, and other similar computing devices. The components shown herein, their connections and relationships, and their functions, are meant to be examples only, and are not meant to limit implementations of the disclosure described and/or claimed herein.

As shown in fig. 10, the electronic device 1000 includes a computing unit 1001 that can perform various appropriate actions and processes according to a computer program stored in a Read Only Memory (ROM)1002 or a computer program loaded from a storage unit 1008 into a Random Access Memory (RAM) 1003. In the RAM 1003, various programs and data necessary for the operation of the electronic apparatus 1000 can also be stored. The calculation unit 1001, the ROM 1002, and the RAM 1003 are connected to each other by a bus 1004. An input/output (I/O) interface 1005 is also connected to bus 1004.

A number of components in the electronic device 1000 are connected to the I/O interface 1005, including: an input unit 1006 such as a keyboard, a mouse, and the like; an output unit 1007 such as various types of displays, speakers, and the like; a storage unit 1008 such as a magnetic disk, optical disk, or the like; and a communication unit 1009 such as a network card, a modem, a wireless communication transceiver, or the like. The communication unit 1009 allows the electronic device 1000 to exchange information/data with other devices through a computer network such as the internet and/or various telecommunication networks.

Computing unit 1001 may be a variety of general and/or special purpose processing components with processing and computing capabilities. Some examples of the computing unit 1001 include, but are not limited to, a Central Processing Unit (CPU), a Graphics Processing Unit (GPU), various dedicated Artificial Intelligence (AI) computing chips, various computing units running machine learning model algorithms, a Digital Signal Processor (DSP), and any suitable processor, controller, microcontroller, and so forth. The calculation unit 1001 executes the respective methods and processes described above, such as the living body detection method and the training method of the living body detection model. For example, in some embodiments, the liveness detection method and the training method of the liveness detection model may be implemented as a computer software program tangibly embodied in a machine-readable medium, such as the storage unit 1008. In some embodiments, part or all of the computer program may be loaded and/or installed onto electronic device 1000 via ROM 1002 and/or communications unit 1009. When the computer program is loaded into the RAM 1003 and executed by the computing unit 1001, one or more steps of the above-described living body detection method and training method of the living body detection model may be performed. Alternatively, in other embodiments, the computing unit 1001 may be configured by any other suitable means (e.g. by means of firmware) to perform the liveness detection method and the training method of the liveness detection model.

Various implementations of the systems and techniques described here above may be implemented in digital electronic circuitry, integrated circuitry, Field Programmable Gate Arrays (FPGAs), Application Specific Integrated Circuits (ASICs), Application Specific Standard Products (ASSPs), system on a chip (SOCs), Complex Programmable Logic Devices (CPLDs), computer hardware, firmware, software, and/or combinations thereof. These various embodiments may include: implemented in one or more computer programs that are executable and/or interpretable on a programmable system including at least one programmable processor, which may be special or general purpose, receiving data and instructions from, and transmitting data and instructions to, a storage system, at least one input device, and at least one output device.

Program code for implementing the methods of the present disclosure may be written in any combination of one or more programming languages. These program codes may be provided to a processor or controller of a general purpose computer, special purpose computer, or other programmable data processing apparatus, such that the program codes, when executed by the processor or controller, cause the functions/operations specified in the flowchart and/or block diagram to be performed. The program code may execute entirely on the machine, partly on the machine, as a stand-alone software package partly on the machine and partly on a remote machine or entirely on the remote machine or server.

In the context of this disclosure, a machine-readable medium may be a tangible medium that can contain, or store a program for use by or in connection with an instruction execution system, apparatus, or device. The machine-readable medium may be a machine-readable signal medium or a machine-readable storage medium. A machine-readable medium may include, but is not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any suitable combination of the foregoing. More specific examples of a machine-readable storage medium would include an electrical connection based on one or more wires, a portable computer diskette, a hard disk, a Random Access Memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disc read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.

To provide for interaction with a user, the systems and techniques described here can be implemented on a computer having: a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to a user; and a keyboard and a pointing device (e.g., a mouse or a trackball) by which a user can provide input to the computer. Other kinds of devices may also be used to provide for interaction with a user; for example, feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user may be received in any form, including acoustic, speech, or tactile input.

The systems and techniques described here can be implemented in a computing system that includes a back-end component (e.g., as a data server), or that includes a middleware component (e.g., an application server), or that includes a front-end component (e.g., a user computer having a graphical user interface or a web browser through which a user can interact with an implementation of the systems and techniques described here), or any combination of such back-end, middleware, or front-end components. The components of the system can be interconnected by any form or medium of digital data communication (e.g., a communication network). Examples of communication networks include: local Area Networks (LANs), Wide Area Networks (WANs), and the Internet.

The computer system may include clients and servers. A client and server are generally remote from each other and typically interact through a communication network. The relationship of client and server arises by virtue of computer programs running on the respective computers and having a client-server relationship to each other. The server may be a cloud server, a server of a distributed system, or a server with a combined blockchain.

It should be understood that various forms of the flows shown above may be used, with steps reordered, added, or deleted. For example, the steps described in the present disclosure may be executed in parallel, sequentially, or in different orders, as long as the desired results of the technical solutions disclosed in the present disclosure can be achieved, and the present disclosure is not limited herein.

The above detailed description should not be construed as limiting the scope of the disclosure. It should be understood by those skilled in the art that various modifications, combinations, sub-combinations and substitutions may be made in accordance with design requirements and other factors. Any modification, equivalent replacement, and improvement made within the spirit and principle of the present disclosure should be included in the scope of protection of the present disclosure.

Claims

1. A method of in vivo detection comprising:

acquiring a face region image and an extended region image of a target object, wherein the face region image includes an image of a first face region, and the extended region image is an image of a second face region and a predetermined extended region other than the second face region;

obtaining target image data according to the face region image and the extended region image, wherein the target image data comprises at least one of the following: a first fused image and auxiliary image data, the auxiliary image data comprising at least one of: a contour image and a second fused image; and

and performing living body detection by using the target image data to obtain a living body detection result aiming at the target object.

2. The method of claim 1, wherein said deriving target image data from the face region image and the extended region image comprises:

splicing the face area image and the extended area image to obtain a first fusion image;

obtaining the auxiliary image data from at least one of the first fused image, the face region image, and the extended region image; and

and obtaining the target image data according to at least one of the first fusion image and the auxiliary image data.

3. The method of claim 2, wherein said deriving the auxiliary image data from at least one of the first fused image, the face region image, and the extended region image comprises:

processing the first fusion image by using a contour extraction algorithm to obtain a contour image;

obtaining the second fused image using one of:

performing color space conversion on the first fusion image to obtain a second fusion image; and

and obtaining the second fusion image according to a conversion face area image and at least one conversion extension area image, wherein the conversion face area image is obtained by performing color space conversion on the face area image, and the conversion extension area image is obtained by performing color conversion on the extension area image.

4. The method according to any one of claims 1 to 3, wherein the performing the living body detection by using the target image data to obtain a living body detection result for the target object comprises:

performing feature extraction on the target image data to obtain target feature data; and

and performing living body classification according to the target characteristic data to obtain a living body detection result aiming at the target object.

5. The method of any of claims 1-4, further comprising:

extracting an image of the first face region from a scene image of the target object to obtain a face region image; and

and extracting images of the second face area and a preset expansion area except the second face area from the scene image to obtain the expansion area image.

6. A method of training a living body detection model, comprising:

acquiring a sample face area image of a sample object and a sample extended area image, wherein the sample face area image is an image of a first sample face area, and the sample extended area image is an image of a second sample face area and a predetermined sample extended area other than the second sample face area;

obtaining sample image data from the sample face region image and the sample extended region image, wherein the sample image data includes at least one of: a first sample fused image and sample auxiliary image data, the sample auxiliary image data comprising at least one of: the sample outline image and the second sample fusion image; and

and training a deep learning model by using the sample image data to obtain the in-vivo detection model.

7. The method of claim 6, wherein said deriving sample image data from the sample face region image and the sample extended region image comprises:

splicing the sample face area image of the sample object and the sample expansion area image to obtain the first sample fusion image;

obtaining the sample assistant image data from at least one of the first sample fusion image, the sample face region image, and the sample extension region image; and

obtaining the sample image data from at least one of the first sample fusion image and the sample auxiliary image data.

8. The method of claim 6 or 7, wherein the training of a deep learning model using the sample image data to derive the liveness detection model comprises:

inputting the sample image data into the deep learning model to obtain a living body detection result of the sample object;

inputting the in-vivo detection result and the in-vivo detection label value of the sample object into a preset loss function to obtain an output value; and

and adjusting the model parameters of the deep learning model according to the output value until a preset condition is met to obtain the in-vivo detection model.

9. A living body detection apparatus comprising:

a first acquisition module configured to acquire a face region image of a target object and an extended region image, wherein the face region image is an image of a first face region, and the extended region image is an image of a second face region and a predetermined extended region other than the second face region;

a first obtaining module, configured to obtain target image data according to the face region image and the extended region image, where the target image data includes at least one of: a first fused image and auxiliary image data, the auxiliary image data comprising at least one of: a contour image and a second fused image; and

and the second obtaining module is used for carrying out living body detection by using the target image data to obtain a living body detection result aiming at the target object.

10. The apparatus of claim 9, wherein the first obtaining means comprises:

the first obtaining submodule is used for splicing the face region image and the extended region image to obtain a first fusion image;

a second obtaining sub-module configured to obtain the auxiliary image data based on at least one of the first fusion image, the face region image, and the extended region image; and

and the third obtaining submodule is used for obtaining the target image data according to at least one of the first fusion image and the auxiliary image data.

11. The apparatus of claim 10, wherein the second obtaining submodule comprises:

the first obtaining unit is used for processing the first fusion image by utilizing a contour extraction algorithm to obtain a contour image;

obtaining the second fused image using one of:

a second obtaining unit, configured to perform color space conversion on the first fused image to obtain a second fused image; and

a third obtaining unit configured to obtain the second fusion image based on a converted face-region image obtained by color-space converting the face-region image and at least one converted extended-region image obtained by color-converting the extended-region image.

12. The apparatus of any of claims 9-11, wherein the second obtaining module comprises:

the fourth obtaining submodule is used for carrying out feature extraction on the target image data to obtain target feature data; and

13. The apparatus of any of claims 9-12, further comprising:

a third obtaining module, configured to extract an image of the first face region from a scene image of the target object, so as to obtain the face region image; and

a fourth obtaining module, configured to extract, from the scene image, an image of the second face region and a predetermined extended region other than the second face region, to obtain the extended region image.

14. A training apparatus for a living body detection model, comprising:

a second acquisition module configured to acquire a sample face area image of a sample object and a sample extended area image, wherein the sample face area image is an image of a first sample face area, and the sample extended area image is an image of a second sample face area and a predetermined extended area other than the second sample face area;

a fifth obtaining module, configured to obtain sample image data according to the sample face area image and the sample extended area image, where the sample image data includes at least one of: a first sample fused image and sample auxiliary image data, the sample auxiliary image data comprising at least one of: the sample outline image and the second sample fusion image; and

and the sixth obtaining module is used for training a deep learning model by using the sample image data to obtain the in-vivo detection model.

15. The apparatus of claim 14, wherein the fifth obtaining means comprises:

a sixth obtaining submodule, configured to splice the sample face region image and the sample extended region image to obtain the first sample fusion image;

a seventh obtaining sub-module configured to obtain the sample auxiliary image data from at least one of the first sample fusion image, the sample face area image, and the sample extension area image; and

an eighth obtaining sub-module, configured to obtain the sample image data according to at least one of the first sample fusion image and the sample auxiliary image data.

16. The apparatus of claim 14 or 15, wherein the sixth obtaining means comprises:

a ninth obtaining submodule, configured to input the sample image data into the deep learning model, and obtain a living body detection result of the sample object;

a tenth obtaining submodule, configured to input the in-vivo detection result of the sample object and the in-vivo detection tag value into a predetermined loss function, and obtain an output value; and

and the eleventh obtaining submodule is used for adjusting the model parameters of the deep learning model according to the output value until a preset condition is met, so that the living body detection model is obtained.

17. An electronic device, comprising:

at least one processor; and

a memory communicatively coupled to the at least one processor; wherein the content of the first and second substances,

the memory stores instructions executable by the at least one processor to enable the at least one processor to perform the method of any one of claims 1 to 5 or any one of claims 6 to 8.

18. A non-transitory computer readable storage medium having stored thereon computer instructions for causing a computer to perform the method of any of claims 1-5 or any of claims 6-8.

19. A computer program product comprising a computer program which, when executed by a processor, implements the method of any of claims 1 to 5 or any of claims 6 to 8.