Face database update method, device, electronic device and storage medium

Abstract

The present invention provides a method, apparatus, electronic device, and storage medium for updating a face database. The method includes: determining the image weight corresponding to each image in a target image set based on the acquisition time of the images in the target image set and a preset offset; the target image set includes a to-be-screened image set and face database images; extracting a feature vector set corresponding to the target image set, and determining an image similarity matrix based on the feature vector set; determining a feature score corresponding to each image in the target image set based on the image weight and the image similarity matrix; determining preferred images based on the feature score corresponding to each image and a preset similarity threshold, and updating the face database based on the preferred images. The method, apparatus, electronic device, and storage medium for updating a face database provided by the present invention can solve the defect of the difficult update of the face recognition database in the prior art, and achieve the reduction of the difficulty in updating the face recognition database.

Description

Technical Field

[0001] The present invention relates to the technical field of image processing, and particularly to a method, device, electronic device and storage medium for updating a face database. Background Art

[0002] With the development of computer vision technology, visual biometric technologies represented by face recognition have developed rapidly. Face recognition refers to a biometric technology that uniquely identifies or verifies a person by comparing and analyzing patterns based on an individual's facial contours. The emergence and wide application of deep learning technology have enabled the booming development of the field of artificial intelligence. As an important branch of it, face recognition has a wide range of applications in our actual life and work.

[0003] The application of face recognition technology is becoming more and more extensive, and face recognition algorithms are becoming more and more accurate. However, with the passage of time, people's appearances will change greatly, including the influence of on-site scene changes such as light on recognition, which will lead to a significant increase in the false recognition rate and rejection rate of face recognition algorithms after a period of use. At this time, it is necessary to spend a lot of time and effort to re-collect face pictures and update the face recognition database to effectively solve this problem, and the difficulty of such operation is very high. Summary of the Invention

[0004] The present invention provides a method, device, electronic device and storage medium for updating a face database, so as to solve the defect of high difficulty in updating the face recognition database in the prior art, and realize reducing the difficulty of updating the face recognition database.

[0005] The present invention provides a method for updating a face database, including:

[0006] Determining the picture weight corresponding to each picture in the target picture set based on the picture acquisition time and a preset offset in the target picture set; wherein, the target picture set includes a picture set to be screened and face database pictures;

[0007] Extracting a feature vector set corresponding to the target picture set, and determining a picture similarity matrix based on the feature vector set;

[0008] Determining the feature score corresponding to each picture in the target picture set based on the picture weight and the picture similarity matrix;

[0009] Determining preferred pictures based on the feature score corresponding to each picture and a preset similarity threshold, and updating the face database based on the preferred pictures.

[0010] According to the improved method for updating a face database, the determining the picture weight corresponding to each picture in the target picture set based on the picture acquisition time and a preset offset in the target picture set includes:

[0011] Determine the image weight corresponding to each image in the target image set based on the following formula:

[0012] w i = b -(k*i) + ot

[0013] where w i is the weight corresponding to the i-th image in the target image set, b is the base parameter, k is the index parameter, and ot is the preset offset.

[0014] According to the improved face database update method, the base parameter satisfies 1 < b < 3, the index parameter satisfies 1 ≤ k < 2, and the preset offset satisfies 0 < ot < 0.8.

[0015] According to the improved face database update method, determining the feature score corresponding to each image in the target image set based on the image weight and the image similarity matrix includes:

[0016] Based on the image weight and the image similarity matrix, obtain the weighted average similarity of each image in the target image set with other images, and based on the weighted average similarity, obtain the feature score corresponding to each image in the target image set.

[0017] According to the improved face database update method, determining the preferred image based on the feature score corresponding to each image and a preset similarity threshold includes:

[0018] Determine the highest feature score based on the feature score corresponding to each image;

[0019] In the case where the highest feature score is greater than the preset similarity threshold, use the image corresponding to the highest feature score as the preferred image.

[0020] According to the improved face database update method, determining the preferred image based on the feature score corresponding to each image and a preset similarity threshold further includes:

[0021] In the case where the highest feature score is not greater than the preset similarity threshold, use the face database image as the preferred image.

[0022] The present invention also provides a face database update device, including:

[0023] A weight setting module, configured to determine the image weight corresponding to each image in the target image set based on the acquisition time of the images in the target image set and a preset offset; wherein, the target image set includes an image set to be screened and face database images;

[0024] A similarity determination module, configured to extract a feature vector set corresponding to the target picture set, and determine a picture similarity matrix based on the feature vector set;

[0025] A feature score determination module, configured to determine a feature score corresponding to each picture in the target picture set based on the picture weight and the picture similarity matrix;

[0026] An update module, configured to determine preferred pictures based on the feature scores corresponding to each picture and a preset similarity threshold, and update the face database based on the preferred pictures.

[0027] The present invention also provides an electronic device, including a memory, a processor, and a computer program stored on the memory and executable on the processor. When the processor executes the program, the steps of the face database update method described in any one of the above are implemented.

[0028] The present invention also provides a non-transitory computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, the steps of the face database update method described in any one of the above are implemented.

[0029] The present invention also provides a computer program product, including a computer program. When the computer program is executed by a processor, the steps of the face database update method described in any one of the above are implemented.

[0030] The face database update method, device, electronic device, and storage medium provided by the present invention determine the picture weight based on the acquisition time of the pictures in the target picture set, and then combine the picture similarity matrix determined by the feature vector set of the pictures to determine the feature score of each picture. Finally, the obtained feature scores are compared with a preset similarity threshold to determine the preferred pictures and dynamically update the face database. Even if the human face changes over time, face recognition can still be performed based on the real-time and dynamically updated face database. The present invention can solve the defect of the large difficulty in updating the face recognition database in the prior art, and achieve the reduction of the difficulty in updating the face recognition database and the improvement of the face recognition ability. Description of the Drawings

[0031] In order to more clearly illustrate the technical solutions in the present invention or the prior art, the following will briefly introduce the drawings required for the description of the embodiments or the prior art. Obviously, the drawings in the following description are some embodiments of the present invention. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0032] Figure 1It is one of the schematic flowcharts of the face database update method provided by the present invention;

[0033] Figure 2 It is the second of the schematic flowcharts of the face database update method provided by the present invention;

[0034] Figure 3 It is the schematic structural diagram of the face database update device provided by the present invention;

[0035] Figure 4 It is the schematic structural diagram of the electronic device provided by the present invention. Detailed implementation manners

[0036] To make the objectives, technical solutions and advantages of the present invention clearer, the technical solutions in the present invention will be clearly and completely described below with reference to the accompanying drawings in the present invention. Apparently, the described embodiments are some, but not all, of the embodiments of the present invention. All other embodiments obtained by those of ordinary skill in the art without making creative efforts based on the embodiments in the present invention belong to the scope of protection of the present invention.

[0037] Below in conjunction with Figures 1 - 4 Describe the face database update method, device, electronic device and storage medium of the present invention.

[0038] As Figure 1 shown, a face database update method provided by the present invention is characterized by including:

[0039] Step 110: Determine the picture weight corresponding to each picture in the target picture set based on the picture acquisition time and the preset offset in the target picture set; wherein, the target picture set includes a picture set to be screened and face database pictures.

[0040] It can be understood that, since the changes of a human face over time are relatively large, it is necessary to determine the picture most similar to other pictures, that is, the central picture, from recent pictures. Therefore, it is necessary to determine the picture weight corresponding to each picture in the target picture set based on the picture acquisition time.

[0041] Step 120: Extract the feature vector set corresponding to the target picture set, and determine the picture similarity matrix based on the feature vector set.

[0042] It can be understood that, in the face database update method provided by the present invention, the face recognition principle is to generate a feature vector by extracting face features, and then calculate the cosine distance corresponding to the feature vector as the similarity. Therefore, the similarity of a human face can be regarded as the cosine distance of the feature vector.

[0043] Referring to the k-means clustering algorithm to calculate the class center, the center similar to a spatial point is obtained by calculating the average value of all feature vectors in the same class. In the present invention, by calculating the relative cosine distance between all picture feature vectors, the picture with the smallest relative cosine distance corresponding to all picture feature vectors is used as the central picture.

[0044] Step 130: Based on the picture weights and the picture similarity matrix, determine the feature score corresponding to each picture in the target picture set.

[0045] It can be understood that, for example, 100 recent face pictures are selected, compared with each other one by one to obtain a similarity matrix, and then the face pictures are scored according to the time weight, and the face picture with the highest score is selected as the central picture.

[0046] Step 140: Based on the feature score corresponding to each picture and a preset similarity threshold, determine the preferred pictures, and update the face database based on the preferred pictures.

[0047] It can be understood that when the face database needs to be updated, scan the database folder to obtain the id (i.e., identity identifier) of all people, and then traverse the id and call the update by passing in the personal id.

[0048] In some embodiments, as Figure 2 shown, the feature files of the latest 99 on-site face pictures and the feature file of 1 face picture in the face database are read from the on-site face picture folder through the id of the individual. If there are less than 99 on-site face pictures, all of them are taken out. Since there are cases of poor quality or blurriness in the on-site face pictures taken, it is easy to affect the quality of picture update and even cause the final central picture to be blurry. To prevent this situation, a similarity threshold needs to be set to skip those blurry face pictures, that is, when the feature score corresponding to the face picture is lower than the similarity threshold, the face picture is skipped and not read.

[0049] In some embodiments, the determining the picture weight corresponding to each picture in the target picture set based on the picture acquisition time and the preset offset in the target picture set includes:

[0050] Determine the picture weight corresponding to each picture in the target picture set based on the following formula:

[0051] w i =b -(k*i) +ot

[0052] where w i is the weight corresponding to the i-th picture in the target picture set, b is the basic parameter, k is the index parameter, and ot is the preset offset.

[0053] In some embodiments, the basic parameter satisfies 1 < b < 3, the index parameter satisfies 1 ≤ k < 2, and the preset offset satisfies 0 < ot < 0.8. For example, b can be 2, k can be 1, and ot can be 0.5.

[0054] It can be understood that the closer the shooting time of the face image is to the current state of the person, the stronger the timeliness. Therefore, a time weight is considered to be added, and the corresponding time weight function is as follows:

[0055] f(i) = b -(k*i) i = 0, 1, 2…

[0056] Among them, the obtained face pictures are sorted according to the acquisition time, and i represents the acquisition order of the face pictures.

[0057] Based on the above time weight function, the weight influence of the face pictures after the 10th one will be too low. In order to prevent the weights of the last dozens of pictures from being too low, an offset needs to be added to reduce the k value and slow down the weight decay, and the weight array is as follows:

[0058]

[0059] w i = b -(k*i) + ot i = 0, 1, 2…

[0060] Among them, n is the total number of pictures in the target picture set, b = 2, k = -1, and ot = 0.5.

[0061] Since multiple pictures ranked behind approach 0 after exponential operation, the offset ot has a relatively large influence on the weights of the subsequent pictures and also has a relatively large influence on the calculation effect of the central pictures. Therefore, it is necessary to determine the specific value of the offset through the test results corresponding to a large number of central pictures.

[0062] Furthermore, the earlier the acquisition time of the face picture, that is, the larger the acquisition time span, the smaller the corresponding offset value. Specifically, it is possible to test through multiple historical face images, and then determine the corresponding relationship between the acquisition time of the face image and the offset value, and establish a corresponding relationship table, so that when determining the weight of a new face picture later, the corresponding relationship table can be referred to determine the offset corresponding to the face picture.

[0063] Take out the feature files of all the pictures in the target picture set one by one, save them as an array, and then compare each feature file with other files one by one. Traverse each picture in the array, compare the traversed picture with each other picture of this person, and establish a two-dimensional array matrix si for this person to store the comparison results, obtain the similarity between this picture and other pictures, and finally obtain the corresponding similarity matrix as follows:

[0064]

[0065] In some embodiments, determining the feature score corresponding to each picture in the target picture set based on the picture weight and the picture similarity matrix includes:

[0066] Based on the picture weight and the picture similarity matrix, obtain the weighted average similarity between each picture and other pictures in the target picture set, and based on the weighted average similarity, obtain the feature score corresponding to each picture in the target picture set.

[0067] It can be understood that by taking the weighted average of the similarity corresponding to each picture in the similarity matrix and the similarity of other pictures, that is, multiplying the similarity of each picture with other pictures by the weight of other pictures and then summing, and then dividing by the sum of the weights of other pictures, the feature score scores of each feature are obtained as follows:

[0068]

[0069] In some embodiments, determining the preferred picture based on the feature score corresponding to each picture and a preset similarity threshold includes:

[0070] Based on the feature score corresponding to each picture, determine the highest feature score;

[0071] In the case where the highest feature score is greater than the preset similarity threshold, use the picture corresponding to the highest feature score as the preferred picture.

[0072] It can be understood that compare the highest feature score among the feature scores corresponding to all the pictures in the target picture set with the preset similarity threshold. If the highest feature score is greater than the preset similarity threshold, then use the picture corresponding to the highest feature score as the preferred picture and store it in the face database, and store the corresponding feature file in the face database.

[0073] In some embodiments, determining the preferred picture based on the feature score corresponding to each picture and a preset similarity threshold further includes:

[0074] When the highest feature score is not greater than a preset similarity threshold, the face database picture is used as the preferred picture.

[0075] It can be understood that if the highest feature score is not greater than the preset similarity threshold, it is considered that the face picture in the database is the best, and the face picture in the database can be directly used for face recognition.

[0076] In summary, the face database update method provided by the present invention includes: determining the picture weight corresponding to each picture in the target picture set based on the picture acquisition time and preset offset in the target picture set; wherein the target picture set includes a picture set to be screened and face database pictures; extracting the feature vector set corresponding to the target picture set, and determining the picture similarity matrix based on the feature vector set; determining the feature score corresponding to each picture in the target picture set based on the picture weight and the picture similarity matrix; determining the preferred picture based on the feature score corresponding to each picture and a preset similarity threshold, and updating the face database based on the preferred picture.

[0077] In the face database update method provided by the present invention, after determining the picture weight based on the picture acquisition time in the target picture set, and then combining the picture similarity matrix determined by the feature vector set of the picture, the feature score of each picture is determined. Finally, the obtained feature score is compared with the preset similarity threshold to determine the preferred picture and dynamically update the face database. Even if the face changes over time, face recognition can still be performed based on the real-time and dynamically updated face database. The present invention can solve the defect of the large difficulty in updating the face recognition database in the prior art, and achieve the reduction of the difficulty in updating the face recognition database and the improvement of the face recognition ability.

[0078] Next, the face database update device provided by the present invention will be described. The face database update device described below can be correspondingly referred to the face database update method described above.

[0079] As Figure 3 shown, the face database update device 300 provided by the present invention includes: a weight setting module 310, a similarity determination module 320, a feature score determination module 330, and an update module 340.

[0080] The weight setting module 310 is configured to determine the picture weight corresponding to each picture in the target picture set based on the picture acquisition time and preset offset in the target picture set; wherein the target picture set includes a picture set to be screened and face database pictures;

[0081] A similarity determination module 320, configured to extract a feature vector set corresponding to the target picture set, and determine a picture similarity matrix based on the feature vector set;

[0082] A feature score determination module 330, configured to determine a feature score corresponding to each picture in the target picture set based on the picture weight and the picture similarity matrix;

[0083] An update module 340, configured to determine preferred pictures based on the feature score corresponding to each picture and a preset similarity threshold, and update the face database based on the preferred pictures.

[0084] In some embodiments, the weight setting module 310 is further configured to determine the picture weight corresponding to each picture in the target picture set based on the following formula:

[0085] w i =b -(k*i) +ot

[0086] where w i is the weight corresponding to the i-th picture in the target picture set, b is a basic parameter, k is an index parameter, and ot is a preset offset.

[0087] In some embodiments, the basic parameter satisfies 1 < b < 3, the index parameter satisfies 1 ≤ k < 2, and the preset offset satisfies 0 < ot < 0.8.

[0088] In some embodiments, the feature score determination module 330 is further configured to obtain a weighted average similarity of each picture in the target picture set with other pictures based on the picture weight and the picture similarity matrix, and obtain a feature score corresponding to each picture in the target picture set based on the weighted average similarity.

[0089] In some embodiments, the update module 340 includes: a highest score determination unit and a first preference determination unit.

[0090] The highest score determination unit is configured to determine the highest feature score based on the feature score corresponding to each picture;

[0091] The first preference determination unit is configured to use the picture corresponding to the highest feature score as the preferred picture when the highest feature score is greater than the preset similarity threshold.

[0092] In some embodiments, the update module 340 further includes: a second preference determination unit.

[0093] The second preferred determination unit is configured to use the face database picture as the preferred picture when the highest feature score is not greater than a preset similarity threshold.

[0094] The electronic device, computer program product, and storage medium provided by the present invention will be described below. The electronic device, computer program product, and storage medium described below can be correspondingly referred to the face database update method described above.

[0095] Figure 4 An entity structure diagram of an electronic device is illustrated, as Figure 4 shown. The electronic device may include: a processor 410, a communication interface 420, a memory 430, and a communication bus 440. Among them, the processor 410, the communication interface 420, and the memory 430 communicate with each other through the communication bus 440. The processor 410 can call the logical instructions in the memory 430 to execute the face database update method, which includes:

[0096] Step 110: Determine the picture weight corresponding to each picture in the target picture set based on the picture acquisition time and a preset offset in the target picture set; wherein the target picture set includes a picture set to be screened and face database pictures;

[0097] Step 120: Extract the feature vector set corresponding to the target picture set, and determine the picture similarity matrix based on the feature vector set;

[0098] Step 130: Determine the feature score corresponding to each picture in the target picture set based on the picture weight and the picture similarity matrix;

[0099] Step 140: Determine the preferred picture based on the feature score corresponding to each picture and a preset similarity threshold, and update the face database based on the preferred picture.

[0100] In addition, when the logical instructions in the above-mentioned memory 430 are implemented in the form of software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or a part of this technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions for causing a computer device (which can be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods described in various embodiments of the present invention. The aforementioned storage medium includes: various media such as USB flash drives, mobile hard disks, read-only memories (ROM, Read-Only Memory), random access memories (RAM, Random Access Memory), magnetic disks, or optical discs that can store program codes.

[0101] On the other hand, the present invention also provides a computer program product. The computer program product includes a computer program that can be stored on a non-transitory computer-readable storage medium. When the computer program is executed by a processor, the computer can execute the face database update method provided by the above-mentioned various methods. The method includes:

[0102] Step 110: Based on the acquisition time of the pictures in the target picture set and a preset offset, determine the picture weight corresponding to each picture in the target picture set; wherein, the target picture set includes a picture set to be screened and face database pictures;

[0103] Step 120: Extract the feature vector set corresponding to the target picture set, and based on the feature vector set, determine a picture similarity matrix;

[0104] Step 130: Based on the picture weight and the picture similarity matrix, determine the feature score corresponding to each picture in the target picture set;

[0105] Step 140: Based on the feature score corresponding to each picture and a preset similarity threshold, determine the preferred pictures, and update the face database based on the preferred pictures.

[0106] On another aspect, the present invention also provides a non-transitory computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, it is implemented to execute the face database update method provided by the above-mentioned various methods. The method includes:

[0107] Step 110: Based on the acquisition time of the pictures in the target picture set and a preset offset, determine the picture weight corresponding to each picture in the target picture set; wherein, the target picture set includes a picture set to be screened and face database pictures;

[0108] Step 120: Extract the feature vector set corresponding to the target picture set, and determine the picture similarity matrix based on the feature vector set;

[0109] Step 130: Determine the feature score corresponding to each picture in the target picture set based on the picture weight and the picture similarity matrix;

[0110] Step 140: Determine the preferred pictures based on the feature scores corresponding to each picture and a preset similarity threshold, and update the face database based on the preferred pictures.

[0111] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed to multiple network units. Some or all of the modules can be selected according to actual needs to achieve the purpose of the solution of this embodiment. Those of ordinary skill in the art can understand and implement it without creative efforts.

[0112] Through the description of the above embodiments, those skilled in the art can clearly understand that each embodiment can be implemented by means of software plus a necessary general hardware platform, and of course, also by hardware. Based on this understanding, the above technical solution, in essence, or the part that contributes to the prior art can be embodied in the form of a software product, which can be stored in a computer-readable storage medium, such as ROM / RAM, magnetic disk, optical disk, etc., including several instructions for causing a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods described in each embodiment or some parts of the embodiments.

[0113] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that they can still modify the technical solutions recorded in the foregoing embodiments, or perform equivalent replacements for some of the technical features; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for updating a face database, characterized in that Including: Based on the acquisition time of the pictures in the target picture set and a preset offset, determine the picture weight corresponding to each picture in the target picture set, where this includes: Determine the picture weight corresponding to each picture in the target picture set based on the following formula: w i = b -(k*i) + ot where w i is the weight corresponding to the i-th picture in the target picture set, b is a basic parameter, k is an index parameter, and ot is a preset offset; wherein, the target picture set includes a picture set to be screened and face database pictures. Extract the feature vector set corresponding to the target picture set, and based on the feature vector set, determine the picture similarity matrix; Based on the picture weight and the picture similarity matrix, determine the feature score corresponding to each picture in the target picture set, where this includes: Based on the picture weight and the picture similarity matrix, obtain the weighted average similarity of each picture in the target picture set with other pictures, and based on the weighted average similarity, obtain the feature score corresponding to each picture in the target picture set; Based on the feature score corresponding to each picture and a preset similarity threshold, determine the preferred pictures, and update the face database based on the preferred pictures, where this includes: Based on the feature score corresponding to each picture, determine the highest feature score; In the case where the highest feature score is greater than the preset similarity threshold, use the picture corresponding to the highest feature score as the preferred picture.

2. The method for updating a face database according to claim 1, wherein The basic parameter satisfies 1 < b < 3, the index parameter satisfies 1 ≤ k < 2, and the preset offset satisfies 0 < ot < 0.

8.

3. The method for updating a face database according to claim 1, wherein The step of determining the preferred pictures based on the feature score corresponding to each picture and a preset similarity threshold further includes: In the case where the highest feature score is not greater than the preset similarity threshold, use the pictures in the face database as the preferred pictures.

4. A face database update device, characterized in that, Including: A weight setting module, configured to determine the picture weight corresponding to each picture in the target picture set based on the acquisition time of the pictures in the target picture set and a preset offset; where the target picture set includes a picture set to be screened and pictures in the face database; The weight setting module is further configured to determine the picture weight corresponding to each picture in the target picture set based on the following formula: w i = b -(k*i) + ot where w i is the weight corresponding to the i-th picture in the target picture set, b is the basic parameter, k is the index parameter, and ot is the preset offset; A similarity determination module, configured to extract the feature vector set corresponding to the target picture set, and based on the feature vector set, determine the picture similarity matrix; A feature score determination module, configured to determine the feature score corresponding to each picture in the target picture set based on the picture weight and the picture similarity matrix; The feature score determination module is further configured to obtain the weighted average similarity of each picture in the target picture set with other pictures based on the picture weight and the picture similarity matrix, and based on the weighted average similarity, obtain the feature score corresponding to each picture in the target picture set; An update module, configured to determine the preferred pictures based on the feature score corresponding to each picture and a preset similarity threshold, and update the face database based on the preferred pictures; The update module includes: a highest score determination unit and a first preference determination unit; The highest score determination unit is configured to determine the highest feature score based on the feature score corresponding to each picture; The first preferred determination unit is configured to use the picture corresponding to the highest feature score as the preferred picture when the highest feature score is greater than the preset similarity threshold.

5. An electronic device, comprising a memory, a processor, and a computer program stored on the memory and executable on the processor, characterized in that, When the processor executes the program, the steps of the face database update method according to any one of claims 1 to 3 are implemented.

6. A non-transitory computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, the steps of the face database update method according to any one of claims 1 to 3 are implemented.

7. A computer program product comprising a computer program, characterized in that, When the computer program is executed by a processor, the steps of the face database update method according to any one of claims 1 to 3 are implemented.

Images