Face image evaluation method and system, server and computer-readable storage medium

Abstract

The present invention provides a facial image evaluation method and system, a server and a computer-readable storage medium, wherein a facial image evaluation method includes: obtaining a facial image and determining the attribute information corresponding to the facial image; extracting the feature information of the facial image and determining the norm value of the feature information; and evaluating the facial image according to the attribute information and the norm value. By evaluating the facial image according to the attribute information corresponding to the facial image and the norm value corresponding to the feature information of the facial image, the quality of the facial image is evaluated from multiple different dimensions, and a comprehensive scoring of the quality of the facial image is achieved, so that a relatively good facial image can be selected in the video stream for facial recognition, and the facial recognition pass rate is guaranteed with a standard, thereby improving the user experience.

Description

Technical Field

[0001] The present invention relates to the technical field of face image evaluation, and in particular to a face image evaluation method, a face image evaluation system, a server and a computer-readable storage medium. Background Art

[0002] In related technologies, scenarios such as online car-hailing driver identity verification, finance, and access control require the application of face recognition technology for identity recognition. Generally, most face recognition scenarios use cameras to collect user face video streams, and then extract a frame containing a face from the video stream for face recognition. However, during the video collection process, users are often in a "semi-cooperative" state: the user's head posture may be adjusted during the collection process (side face, lowered head), the face may be blurred during movement, and the face may be blocked occasionally, etc., all of which will affect the accuracy of subsequent face recognition. How to select the face picture with the best face recognition effect from the video stream is crucial to ensuring the pass rate of face recognition and improving user experience. Summary of the invention

[0003] The present invention aims to solve at least one of the technical problems existing in the prior art or related art.

[0004] To this end, a first aspect of the present invention provides a facial image evaluation method.

[0005] A second aspect of the present invention provides a facial image evaluation system.

[0006] A third aspect of the present invention provides a server.

[0007] A fourth aspect of the present invention provides a computer-readable storage medium.

[0008] In view of this, the first aspect of the present invention provides a facial image evaluation method, comprising: acquiring a facial image and determining attribute information corresponding to the facial image; extracting feature information of the facial image and determining the norm value of the feature information; and evaluating the facial image based on the attribute information and the norm value.

[0009] In this technical solution, the facial image is evaluated based on the attribute information corresponding to the facial image and the norm value corresponding to the facial image feature information, and then the quality of the facial image is evaluated from multiple different dimensions to achieve a comprehensive scoring of the facial image quality. Relatively better facial images can be selected in the video stream for facial recognition. The target ensures the facial recognition pass rate and improves the user experience.

[0010] In addition, the face image evaluation method in the above technical solution provided by the present invention may also have the following additional technical features:

[0011] In the above technical solution, the step of evaluating the facial image according to the attribute information and the norm value specifically includes: determining a first score value corresponding to the attribute information; obtaining a test set, and determining the weight value corresponding to the attribute information and the norm value in the test set through a Bayesian algorithm; calculating a second score value of the facial image according to the first score value, the norm value and the corresponding weight value, so as to evaluate the facial image through the second score value.

[0012] In this technical solution, when evaluating a face image according to the attribute information and the norm value, a first score value is first calculated according to the attribute information marked in the face image. Specifically, the process of calculating the first score value can be automatically calculated by a trained convolution algorithm. Taking the attribute information including the proportion of face occlusion as an example, the greater the proportion of face occlusion, the lower the success rate of face recognition, and thus the lower the first score value. That is, the first score value can positively reflect the success rate of face recognition.

[0013] After calculating the first score value, a test set is further obtained, and the weight value corresponding to each attribute information in the test set is determined by the Bayesian algorithm. The weight value reflects the importance of each attribute information in the multiple attribute information. Taking the attribute information including the proportion of face occlusion and the light intensity of the face image as an example, among the above two types of attribute information, the impact of face occlusion on the success rate of face recognition is greater than the impact of light intensity on the success rate of face recognition. Therefore, the corresponding weight value of the proportion of face occlusion is relatively greater than the weight value of light intensity.

[0014] After determining the first score value and the corresponding weight value, the product of the first score value and the norm value and the corresponding weight value is calculated to obtain the second score value of the face image, that is, the score value of the face image calculated according to the multiple attribute information and the introduced norm value concept. The face image can be accurately evaluated according to the second score value, thereby ensuring the accuracy of face recognition.

[0015] In any of the above technical solutions, before the step of obtaining the test set, the face image evaluation method also includes: obtaining an initial image and a face recognition model, and marking attribute information in the initial image to obtain a training image set; training the face recognition model through the training image set to obtain a prediction model and a training set, and determining the attribute information corresponding to the face image through the prediction model.

[0016] In this technical solution, the face recognition model is trained so that the face recognition model can simultaneously predict the attribute information and norm value of the face image. First, face image data of various postures, scenes and races are prepared as initial images, wherein the initial image is an RGB (a color specification) image.

[0017] In the initial image, multiple attribute information of the face in the image is annotated, such as posture, blur level, occlusion and light intensity, and the annotated initial image is used to form a training image set. The face recognition model is trained using the training image set to obtain a trained prediction model and training set, and the weight values ​​of multiple attribute information are determined through the training set, and multiple attribute values ​​of face information are predicted through the prediction model.

[0018] Among them, the face recognition model can be trained through multi-task training.

[0019] In any of the above technical solutions, the attribute information includes at least one of the following: the posture of the face, the blurriness of the face, the proportion of the face that is blocked, and the light intensity of the face image.

[0020] In this technical solution, the posture of the face specifically refers to the face facing the camera, or tilting the head, lowering the head, etc. The blurriness of the face is reflected in whether the face is in motion or whether the focus is accurate in the frame image. The proportion of the face that is blocked can reflect whether the entire face can be recognized. The light intensity of the face image is intuitively described as the image brightness or exposure. Too high or too low brightness or exposure may affect the accuracy of face recognition.

[0021] It is understandable that the attribute information is not limited to the above-mentioned attributes, and any image attribute that may affect the success rate of face recognition can be used as a feasible embodiment of the present invention.

[0022] The second aspect of the present invention provides a facial image evaluation system, comprising: an acquisition unit, configured to acquire a facial image and determine attribute information corresponding to the facial image; a processing unit, configured to extract feature information of the facial image and determine a norm value of the feature information; an evaluation unit, configured to evaluate the facial image based on the attribute information and the norm value.

[0023] In this technical solution, the facial image is evaluated based on the attribute information corresponding to the facial image and the norm value corresponding to the facial image feature information, and then the quality of the facial image is evaluated from multiple different dimensions to achieve a comprehensive scoring of the facial image quality. Relatively better facial images can be selected in the video stream for facial recognition. The target ensures the facial recognition pass rate and improves the user experience.

[0024] In the above technical solution, the evaluation unit is specifically used to: determine a first score value corresponding to the attribute information; obtain a test set, and determine the weight value corresponding to the attribute information and the norm value in the test set through a Bayesian algorithm; calculate the second score value of the face image based on the first score value, the norm value and the corresponding weight value, so as to evaluate the face image through the second score value.

[0025] In this technical solution, when evaluating a face image according to the attribute information and the norm value, a first score value is first calculated according to the attribute information marked in the face image. Specifically, the process of calculating the first score value can be automatically calculated by a trained convolution algorithm. Taking the attribute information including the proportion of face occlusion as an example, the greater the proportion of face occlusion, the lower the success rate of face recognition, and thus the lower the first score value. That is, the first score value can positively reflect the success rate of face recognition.

[0026] After calculating the first score value, a test set is further obtained, and the weight value corresponding to each attribute information in the test set is determined by the Bayesian algorithm. The weight value reflects the importance of each attribute information in the multiple attribute information. Taking the attribute information including the proportion of face occlusion and the light intensity of the face image as an example, among the above two types of attribute information, the impact of face occlusion on the success rate of face recognition is greater than the impact of light intensity on the success rate of face recognition. Therefore, the corresponding weight value of the proportion of face occlusion is relatively greater than the weight value of light intensity.

[0027] After determining the first score value and the corresponding weight value, the product of the first score value and the norm value and the corresponding weight value is calculated to obtain the second score value of the face image, that is, the score value of the face image calculated according to the multiple attribute information and the introduced norm value concept. The face image can be accurately evaluated according to the second score value, thereby ensuring the accuracy of face recognition.

[0028] In any of the above technical solutions, the face image evaluation system also includes: an acquisition unit, configured to acquire an initial image and a face recognition model, and to mark attribute information in the initial image to obtain a training image set; a training unit, configured to train the face recognition model through the training image set to obtain a prediction model and a training set, and to determine the attribute information corresponding to the face image through the prediction model.

[0029] In this technical solution, the face recognition model is trained so that the face recognition model can simultaneously predict the attribute information and norm value of the face image. First, face image data of various postures, scenes and races are prepared as initial images, wherein the initial image is an RGB (a color specification) image.

[0030] In the initial image, multiple attribute information of the face in the image is annotated, such as posture, blur level, occlusion and light intensity, and the annotated initial image is used to form a training image set. The face recognition model is trained using the training image set to obtain a trained prediction model and training set, and the weight values ​​of multiple attribute information are determined through the training set, and multiple attribute values ​​of face information are predicted through the prediction model.

[0031] Among them, the face recognition model can be trained through multi-task training.

[0032] In any of the above technical solutions, the attribute information includes at least one of the following: the posture of the face, the blurriness of the face, the proportion of the face that is blocked, and the light intensity of the face image.

[0033] In this technical solution, the posture of the face specifically refers to the face facing the camera, or tilting the head, lowering the head, etc. The blurriness of the face is reflected in whether the face is in motion or whether the focus is accurate in the frame image. The proportion of the face that is blocked can reflect whether the entire face can be recognized. The light intensity of the face image is intuitively described as the image brightness or exposure. Too high or too low brightness or exposure may affect the accuracy of face recognition.

[0034] It is understandable that the attribute information is not limited to the above-mentioned attributes, and any image attribute that may affect the success rate of face recognition can be used as a feasible embodiment of the present invention.

[0035] The third aspect of the present invention provides a server, comprising: a memory configured to store a computer program; a processor configured to execute the computer program to implement a facial image evaluation method as provided in any of the above technical solutions. Therefore, the server includes all the beneficial effects of the facial image evaluation method provided in any of the above technical solutions, which will not be repeated here.

[0036] The fourth aspect of the present invention provides a computer-readable storage medium having a computer program stored thereon. When the computer program is executed by a processor, the facial image evaluation method provided in any of the above-mentioned technical solutions is implemented. Therefore, the computer-readable storage medium includes all the beneficial effects of the facial image evaluation method provided in any of the above-mentioned technical solutions, which will not be repeated here. BRIEF DESCRIPTION OF THE DRAWINGS

[0037] The above and / or additional aspects and advantages of the present invention will become apparent and easily understood from the description of the embodiments in conjunction with the following drawings, in which:

[0038] Figure 1 A flow chart of a method for evaluating a face image according to an embodiment of the present invention is shown;

[0039] Figure 2 Another flow chart of a method for evaluating a face image according to an embodiment of the present invention is shown;

[0040] Figure 3 Another flow chart of a facial image evaluation method according to an embodiment of the present invention is shown;

[0041] Figure 4A structural block diagram of a face image evaluation system according to an embodiment of the present invention is shown;

[0042] Figure 5 Another structural block diagram of a face image evaluation system according to an embodiment of the present invention is shown;

[0043] Figure 6 A structural block diagram of a server according to an embodiment of the present invention is shown. DETAILED DESCRIPTION

[0044] In order to more clearly understand the above-mentioned purpose, features and advantages of the present invention, the present invention is further described in detail below in conjunction with the accompanying drawings and specific embodiments. It should be noted that the embodiments of the present application and the features in the embodiments can be combined with each other without conflict.

[0045] In the following description, many specific details are set forth to facilitate a full understanding of the present invention. However, the present invention may also be implemented in other ways different from those described herein. Therefore, the protection scope of the present invention is not limited to the specific embodiments disclosed below.

[0046] Refer to the following Figures 1 to 6 The face image evaluation method, face image evaluation system, server and computer-readable storage medium according to some embodiments of the present invention are described.

[0047] Embodiment 1

[0048] like Figure 1 As shown, in one embodiment of the present invention, a face image evaluation method is provided, comprising:

[0049] Step S102, obtaining a face image and determining attribute information corresponding to the face image;

[0050] Step S104, extracting feature information of the face image and determining a norm value of the feature information;

[0051] Step S106, evaluating the face image according to the attribute information and the norm value.

[0052] In this embodiment, the facial image is evaluated based on the attribute information corresponding to the facial image and the norm value corresponding to the facial image feature information, and then the quality of the facial image is evaluated from multiple different dimensions to achieve a comprehensive scoring of the facial image quality. Relatively better facial images can be selected in the video stream for facial recognition. The target ensures the facial recognition pass rate and improves the user experience.

[0053] like Figure 2As shown, in one embodiment of the present invention, the step of evaluating the face image according to the attribute information and the norm value specifically includes:

[0054] Step S202, determining a first score value corresponding to the attribute information;

[0055] Step S204, obtaining a test set, and determining weight values ​​corresponding to attribute information and norm values ​​in the test set by using a Bayesian algorithm;

[0056] Step S206, calculating a second score value of the face image according to the first score value, the norm value and the corresponding weight value, so as to evaluate the face image through the second score value.

[0057] In this embodiment, when evaluating the face image according to the attribute information and the norm value, the first score value is first calculated according to the attribute information marked in the face image. Specifically, the process of calculating the first score value can be automatically calculated by a trained convolution algorithm. Taking the attribute information including the proportion of the face being blocked as an example, the greater the proportion of the face being blocked, the lower the success rate of face recognition, and thus the lower the first score value. That is, the first score value can positively reflect the success rate of face recognition.

[0058] After calculating the first score value, a test set is further obtained, and the weight value corresponding to each attribute information in the test set is determined by the Bayesian algorithm. The weight value reflects the importance of each attribute information in the multiple attribute information. Taking the attribute information including the proportion of face occlusion and the light intensity of the face image as an example, among the above two types of attribute information, the impact of face occlusion on the success rate of face recognition is greater than the impact of light intensity on the success rate of face recognition. Therefore, the corresponding weight value of the proportion of face occlusion is relatively greater than the weight value of light intensity.

[0059] After determining the first score value and the corresponding weight value, the product of the first score value and the norm value and the corresponding weight value is calculated to obtain the second score value of the face image, that is, the score value of the face image calculated according to the multiple attribute information and the introduced norm value concept. The face image can be accurately evaluated according to the second score value, thereby ensuring the accuracy of face recognition.

[0060] like Figure 3 As shown, in one embodiment of the present invention, before the step of obtaining the test set, the face image evaluation method further includes:

[0061] Step S302, obtaining an initial image and a face recognition model, and marking attribute information in the initial image to obtain a training image set;

[0062] Step S304: training the face recognition model through the training image set to obtain a prediction model and a training set, and determining attribute information corresponding to the face image through the prediction model.

[0063] In this embodiment, the face recognition model is trained so that the face recognition model can simultaneously predict the attribute information and norm value of the face image. First, face image data of various postures, scenes and races are prepared as initial images, wherein the initial image is an RGB (a color specification) image.

[0064] In the initial image, multiple attribute information of the face in the image is annotated, such as posture, blur level, occlusion and light intensity, and the annotated initial image is used to form a training image set. The face recognition model is trained using the training image set to obtain a trained prediction model and training set, and the weight values ​​of multiple attribute information are determined through the training set, and multiple attribute values ​​of face information are predicted through the prediction model.

[0065] Among them, the face recognition model can be trained through multi-task training.

[0066] In one embodiment of the present invention, the attribute information includes at least one of the following: the posture of the face, the blurriness of the face, the proportion of the face that is blocked, and the illumination intensity of the face image.

[0067] In this embodiment, the posture of the face specifically refers to the face facing the camera, or tilting the head, lowering the head, etc. The blurriness of the face is reflected in whether the face is in motion or whether the focus is accurate in the frame image. The proportion of the face being blocked can reflect whether the entire face can be recognized. The light intensity of the face image is intuitively described as the image brightness or exposure. Too high or too low brightness or exposure may affect the accuracy of face recognition.

[0068] It is understandable that the attribute information is not limited to the above-mentioned attributes, and any image attribute that may affect the success rate of face recognition can be used as a feasible embodiment of the present invention.

[0069] Embodiment 2

[0070] In a complete embodiment of the present invention, a multi-dimensional facial image evaluation method is provided.

[0071] Specifically, the present invention evaluates the quality of facial images from multiple dimensions such as facial posture, lighting, blur level, occlusion, closed eyes, and intense expressions, and uses a facial recognition model to extract facial image features to obtain the facial feature NORM (norm function) value. Adding the dimension of the facial NORM value to evaluate facial quality can directly reflect whether the facial image is "suitable for facial recognition."

[0072] Determine the weight of each dimension: Through Bayesian search, quickly and accurately determine the weight of different dimensions such as face posture, blur, occlusion, lighting, NORM value, etc. that affect face quality.

[0073] The specific process is as follows:

[0074] The first step is face data preparation: collect face data of various postures, scenes, and races.

[0075] The second step is to label the attribute values ​​of each dimension of the face: posture (yaw, pitch), blur level, occlusion, light intensity, etc.

[0076] The third step is to calculate the NORM value of facial features: use the face recognition model to extract the features of the face data and calculate the NORM value of the features as a label.

[0077] The fourth step is to input RGB face images during training and use a multi-task training method. The model simultaneously predicts multiple attributes of the face, such as posture, blur, occlusion, and NORM value.

[0078] The fifth step is to use Bayesian search for the weights of attributes of different dimensions on the test set after training, and the sum of the weights is 1.

[0079] Step 6: Face quality assessment: Use the trained model to predict the scores of different face attributes, and then calculate the comprehensive score of face quality through the weights obtained by 5 searches.

[0080] Embodiment 3

[0081] like Figure 4 As shown, in one embodiment of the present invention, a face image evaluation system 400 is provided, including: an acquisition unit 402, configured to acquire a face image and determine attribute information corresponding to the face image; a processing unit 404, configured to extract feature information of the face image and determine a norm value of the feature information; an evaluation unit 406, configured to evaluate the face image according to the attribute information and the norm value.

[0082] In this embodiment, the facial image is evaluated based on the attribute information corresponding to the facial image and the norm value corresponding to the facial image feature information, and then the quality of the facial image is evaluated from multiple different dimensions to achieve a comprehensive scoring of the facial image quality. Relatively better facial images can be selected in the video stream for facial recognition. The target ensures the facial recognition pass rate and improves the user experience.

[0083] In one embodiment of the present invention, the evaluation unit 406 is specifically used to: determine a first score value corresponding to the attribute information; obtain a test set, and determine the weight value corresponding to the attribute information and the norm value in the test set through a Bayesian algorithm; calculate a second score value of the face image based on the first score value, the norm value and the corresponding weight value, so as to evaluate the face image through the second score value.

[0084] In this embodiment, when evaluating the face image according to the attribute information and the norm value, the first score value is first calculated according to the attribute information marked in the face image. Specifically, the process of calculating the first score value can be automatically calculated by a trained convolution algorithm. Taking the attribute information including the proportion of the face being blocked as an example, the greater the proportion of the face being blocked, the lower the success rate of face recognition, and thus the lower the first score value. That is, the first score value can positively reflect the success rate of face recognition.

[0085] After calculating the first score value, a test set is further obtained, and the weight value corresponding to each attribute information in the test set is determined by the Bayesian algorithm. The weight value reflects the importance of each attribute information in the multiple attribute information. Taking the attribute information including the proportion of face occlusion and the light intensity of the face image as an example, among the above two types of attribute information, the impact of face occlusion on the success rate of face recognition is greater than the impact of light intensity on the success rate of face recognition. Therefore, the corresponding weight value of the proportion of face occlusion is relatively greater than the weight value of light intensity.

[0086] After determining the first score value and the corresponding weight value, the product of the first score value and the norm value and the corresponding weight value is calculated to obtain the second score value of the face image, that is, the score value of the face image calculated according to the multiple attribute information and the introduced norm value concept. The face image can be accurately evaluated according to the second score value, thereby ensuring the accuracy of face recognition.

[0087] like Figure 5 As shown, in one embodiment of the present invention, a face image evaluation system 500 includes: an acquisition unit 502, a processing unit 504, an evaluation unit 506 and a training unit 508, wherein the acquisition unit 502 is configured to acquire an initial image and a face recognition model, and annotate attribute information in the initial image to obtain a training image set; the training unit 508 is configured to train the face recognition model through the training image set to obtain a prediction model and a training set, and determine the attribute information corresponding to the face image through the prediction model.

[0088] In this embodiment, the face recognition model is trained so that the face recognition model can simultaneously predict the attribute information and norm value of the face image. First, face image data of various postures, scenes and races are prepared as initial images, wherein the initial image is an RGB (a color specification) image.

[0089] In the initial image, multiple attribute information of the face in the image is annotated, such as posture, blur level, occlusion and light intensity, and the annotated initial image is used to form a training image set. The face recognition model is trained using the training image set to obtain a trained prediction model and training set, and the weight values ​​of multiple attribute information are determined through the training set, and multiple attribute values ​​of face information are predicted through the prediction model.

[0090] Among them, the face recognition model can be trained through multi-task training.

[0091] In one embodiment of the present invention, the attribute information includes at least one of the following: the posture of the face, the blurriness of the face, the proportion of the face that is blocked, and the illumination intensity of the face image.

[0092] In this embodiment, the posture of the face specifically refers to the face facing the camera, or tilting the head, lowering the head, etc. The blurriness of the face is reflected in whether the face is in motion or whether the focus is accurate in the frame image. The proportion of the face being blocked can reflect whether the entire face can be recognized. The light intensity of the face image is intuitively described as the image brightness or exposure. Too high or too low brightness or exposure may affect the accuracy of face recognition.

[0093] It is understandable that the attribute information is not limited to the above-mentioned attributes, and any image attribute that may affect the success rate of face recognition can be used as a feasible embodiment of the present invention.

[0094] Embodiment 4

[0095] like Figure 6 As shown, in one embodiment of the present invention, a server 600 is provided, including: a memory 602, configured to store a computer program; a processor 604, configured to execute the computer program to implement the face image evaluation method provided in any of the above embodiments. Therefore, the server includes all the beneficial effects of the face image evaluation method provided in any of the above embodiments, which will not be repeated here.

[0096] Embodiment 5

[0097] In one embodiment of the present invention, a computer-readable storage medium is provided, on which a computer program is stored. When the computer program is executed by a processor, the facial image evaluation method provided in any of the above embodiments is implemented. Therefore, the computer-readable storage medium includes all the beneficial effects of the facial image evaluation method provided in any of the above embodiments, which will not be repeated here.

[0098] In the description of the present invention, the term "plurality" refers to two or more than two. Unless otherwise clearly defined, the orientation or positional relationship indicated by the terms "upper" and "lower" is based on the orientation or positional relationship described in the accompanying drawings. It is only for the convenience of describing the present invention and simplifying the description, and does not indicate or imply that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation. Therefore, it cannot be understood as a limitation on the present invention. The terms "connection", "installation", "fixation", etc. should be understood in a broad sense. For example, "connection" can be a fixed connection, a detachable connection, or an integral connection; it can be directly connected or indirectly connected through an intermediate medium. For ordinary technicians in this field, the specific meanings of the above terms in the present invention can be understood according to specific circumstances.

[0099] In the description of the present invention, the description of the terms "one embodiment", "some embodiments", "specific embodiments", etc. means that the specific features, structures, materials or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present invention. In the present invention, the schematic representation of the above terms does not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described can be combined in any one or more embodiments or examples in a suitable manner.

[0100] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. For those skilled in the art, the present invention may have various modifications and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. A facial image evaluation method, characterized in that: include: Acquire a facial image, and determine multiple attribute information corresponding to the facial image; Extracting feature information of the face image and determining a norm value of the feature information; evaluating the facial image according to the attribute information and the norm value; The step of evaluating the face image according to the attribute information and the norm value specifically includes: Determine a plurality of first score values ​​corresponding to the plurality of attribute information; Obtain a test set, and determine weight values ​​corresponding to the attribute information and the norm value in the test set by using a Bayesian algorithm; A second score value of the facial image is calculated according to the first score value, the norm value, and the corresponding weight value, so as to evaluate the facial image through the second score value.

2. The facial image evaluation method according to claim 1, characterized in that: Before the step of obtaining the test set, the face image evaluation method further includes: Acquire an initial image and a face recognition model, and annotate the attribute information in the initial image to obtain a training image set; The face recognition model is trained by using the training image set to obtain a prediction model and a training set, and the attribute information corresponding to the face image is determined by using the prediction model.

3. The facial image evaluation method according to any one of claims 1 to 2, characterized in that: The attribute information includes at least one of the following: The posture of the face, the blur of the face, the proportion of the face that is blocked, and the lighting intensity of the face image.

4. A face image evaluation system, characterized in that: include: An acquisition unit, configured to acquire a face image and determine a plurality of attribute information corresponding to the face image; a processing unit configured to extract feature information of the face image and determine a norm value of the feature information; an evaluation unit, configured to evaluate the face image according to the attribute information and the norm value; The evaluation unit is specifically used for: Determine a plurality of first score values ​​corresponding to the plurality of attribute information; Obtain a test set, and determine weight values ​​corresponding to the attribute information and the norm value in the test set by using a Bayesian algorithm; A second score value of the facial image is calculated according to the first score value, the norm value, and the corresponding weight value, so as to evaluate the facial image through the second score value.

5. The facial image evaluation system according to claim 4, characterized in that: Also includes: The acquisition unit is configured to acquire an initial image and a face recognition model, and mark the attribute information in the initial image to obtain a training image set; The training unit is configured to train the face recognition model through the training image set to obtain a prediction model and a training set, and determine the attribute information corresponding to the face image through the prediction model.

6. The facial image evaluation system according to any one of claims 4 to 5, characterized in that: The attribute information includes at least one of the following: The posture of the face, the blur of the face, the proportion of the face that is blocked, and the lighting intensity of the face image.

7. A server, characterized in that: include: a memory configured to store a computer program; A processor configured to execute the computer program to implement the facial image evaluation method according to any one of claims 1 to 3.

8. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the facial image evaluation method according to any one of claims 1 to 3 is implemented.

Images