CN115393926A - Method and device for improving face recognition precision and server - Google Patents
Method and device for improving face recognition precision and server Download PDFInfo
- Publication number
- CN115393926A CN115393926A CN202210940098.3A CN202210940098A CN115393926A CN 115393926 A CN115393926 A CN 115393926A CN 202210940098 A CN202210940098 A CN 202210940098A CN 115393926 A CN115393926 A CN 115393926A
- Authority
- CN
- China
- Prior art keywords
- face
- face information
- probability
- card
- candidate
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V40/00—Recognition of biometric, human-related or animal-related patterns in image or video data
- G06V40/10—Human or animal bodies, e.g. vehicle occupants or pedestrians; Body parts, e.g. hands
- G06V40/16—Human faces, e.g. facial parts, sketches or expressions
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/70—Arrangements for image or video recognition or understanding using pattern recognition or machine learning
- G06V10/74—Image or video pattern matching; Proximity measures in feature spaces
- G06V10/761—Proximity, similarity or dissimilarity measures
Abstract
The application discloses a method, a device and a server for improving face recognition precision. The method comprises the following steps: acquiring face information of a current card punch, and acquiring candidate face information according to the face information; respectively determining the probability of punching a card at the current terminal equipment, the probability of punching a card at the current time period and the probability of being correctly recognized by the human face corresponding to each candidate human face information according to a historical database; inputting the probability into a Bayesian model to obtain the face matching accuracy of each candidate; and determining the face with the highest face matching accuracy as the target face information. The method and the device can make full use of the existing staff attendance card punching behaviors, match the possible card punching personnel to specific equipment, adopt the data of a recent period of time for the behaviors of the user in the process of predictive analysis, and perform dynamic statistics on the identified terminal equipment, have the function of dynamically updating the data prediction result, improve the accuracy of face recognition and reduce the cost.
Description
Technical Field
The present application relates to the field of face recognition technology, and in particular, to a method, an apparatus, and a server for improving face recognition accuracy.
Background
The existing attendance machine of the face recognition technology for industrial production usually records the picture of a face base in advance, extracts the face characteristics and establishes a face file. The attendance machine acquires the face picture in real time, compares the face picture with the face file, and takes the comparison result as the result of face recognition. However, the prior art has the following problems in practical use:
(1) In the existing factory, the number of face base libraries is large, and the larger the face base library is, the accuracy of recognition is seriously reduced;
(2) In a group, a plurality of factories or departments share one face recognition attendance system, and besides the problem of large quantity of base libraries, pictures of face base libraries uploaded by different departments have different qualities, and the poor quality pictures also influence the accuracy of face comparison.
Disclosure of Invention
The embodiment of the application aims to provide a method, a device and a server for improving face recognition precision, and aims to solve the problems that in the prior art, the number of face base libraries is large, and the quality of uploaded face base library pictures is different, so that the accuracy of face comparison is low.
In order to achieve the above object, a first aspect of the present application provides a method for improving face recognition accuracy, including:
acquiring face information of a current card punch;
acquiring candidate face information according to the face information;
respectively determining the probability of punching a card at the current terminal equipment, the probability of punching a card at the current time period and the probability of being correctly recognized by the human face corresponding to each candidate human face information according to a historical database;
inputting the probability of card punching at the current terminal equipment, the probability of card punching at the current time period and the probability of correct recognition of the face into a Bayesian model to obtain the face matching accuracy of each candidate;
and determining the face with the highest matching accuracy as the target face information.
In the embodiment of the present application, obtaining candidate face information according to face information includes:
and obtaining candidate face information which is most similar to the current card punching face information through a face matching algorithm.
In the embodiment of the application, respectively determining the probability of punching a card at the current terminal device corresponding to each candidate face information according to the historical database, wherein the probability of punching a card at the current time period comprises:
acquiring equipment point data and time data collected by terminal equipment in a historical database;
determining the probability value of each candidate face information on the current terminal equipment card according to the equipment point data;
and determining the probability value of each candidate face information in the current time slot according to the time data in the historical database.
In the embodiment of the application, the determining the probability value of each candidate face information on the current terminal device according to the device point data comprises:
acquiring a code of current terminal equipment;
counting the number of times of punching a card on the current terminal equipment by each candidate face information according to the codes;
and determining the probability value of each candidate face information on the current terminal equipment according to the times of the candidate face information on the current terminal equipment.
In the embodiment of the present application, respectively determining the probability that each candidate face information is corresponding to the face recognition accuracy includes:
acquiring face information input in a face base;
acquiring face information input by terminal equipment;
and calculating the probability of correct face recognition corresponding to each candidate face information according to the face information input in the face base and the face information input by the terminal equipment.
In the embodiment of the application, the probability of being correct by face recognition is represented by the face similarity obtained by a face matching algorithm.
In the embodiment of the application, the Bayesian model satisfies formula (1):
wherein, p (x | s, t, c) is the probability of successful matching when certain face information is on certain terminal equipment in a certain time period and under the condition of acquiring the similarity of matched faces; p (s | x) is the probability of the appearance of the certain face information in the certain face identification terminal equipment; p (t | x) is the probability of the appearance of certain face information in a certain time period; p (c | x) is the probability that certain face information is correct by face recognition; p (x) is a prior probability of using face recognition service for certain face information; p(s) is the probability value of each face information card punching on the terminal equipment; p (t) is the probability value of the possibility of checking attendance and checking the card in a time period of each face information; and p (c) is the face matching similarity.
In an embodiment of the present application, the method further includes:
and updating the historical database according to the target face information, the corresponding terminal equipment and the time.
A second aspect of the present application provides a server comprising:
a memory configured to store instructions; and
a processor configured to call instructions from the memory and upon execution of the instructions to enable a method according to the above for improving the accuracy of face recognition.
A third aspect of the present application provides an apparatus for improving face recognition accuracy, including:
a plurality of terminal devices; and
the server according to the above.
A fourth aspect of the present application provides a machine-readable storage medium having stored thereon instructions for causing a machine to perform the above-described method for improving face recognition accuracy.
By the technical scheme, the face information of the current card punching is obtained, and the candidate face information is obtained according to the face information. And respectively determining the probability of punching a card at the current terminal equipment, the probability of punching a card at the current time period and the probability of being correctly recognized by the face corresponding to each candidate face information according to the historical database. And inputting the probability of punching a card at the current terminal equipment, the probability of punching a card at the current time period and the probability of being correctly recognized by the human face into the Bayesian model to obtain the human face matching accuracy of each candidate. And determining the face with the highest face matching accuracy as the target face information. The method and the device can make full use of the existing staff attendance card punching behaviors, match the possible card punching personnel to specific equipment, adopt the data of a recent period of time for the behaviors of the user in the process of predictive analysis, and perform dynamic statistics on the identified terminal equipment, have the function of dynamically updating the data prediction result, improve the accuracy of face recognition and reduce the cost.
Additional features and advantages of embodiments of the present application will be described in detail in the detailed description which follows.
Drawings
The accompanying drawings, which are included to provide a further understanding of the embodiments of the disclosure and are incorporated in and constitute a part of this specification, illustrate embodiments of the disclosure and together with the description serve to explain the embodiments of the disclosure, but are not intended to limit the embodiments of the disclosure. In the drawings:
fig. 1 schematically illustrates a flow chart of a method for improving face recognition accuracy according to an embodiment of the present application;
FIG. 2 is a block diagram schematically illustrating a server according to an embodiment of the present application;
fig. 3 is a block diagram schematically illustrating an apparatus for improving face recognition accuracy according to an embodiment of the present application.
Description of the reference numerals
310. Multiple terminal device 320 servers
Detailed Description
To make the objects, technical solutions and advantages of the embodiments of the present application clearer, the technical solutions of the embodiments of the present application will be clearly and completely described below with reference to the drawings in the embodiments of the present application, and it should be understood that the specific embodiments described herein are only used for illustrating and explaining the embodiments of the present application and are not used for limiting the embodiments of the present application. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present application.
It should be noted that if directional indications (such as upper, lower, left, right, front, rear, 8230; \8230;) are referred to in the embodiments of the present application, the directional indications are only used for explaining the relative positional relationship between the components in a specific posture (as shown in the attached drawings), the motion situation, etc., and if the specific posture is changed, the directional indications are correspondingly changed.
In addition, if there is a description of "first", "second", etc. in the embodiments of the present application, the description of "first", "second", etc. is for descriptive purposes only and is not to be construed as indicating or implying relative importance or implicitly indicating the number of technical features indicated. Thus, a feature defined as "first" or "second" may explicitly or implicitly include at least one such feature. In addition, technical solutions between various embodiments may be combined with each other, but must be realized by a person skilled in the art, and when the technical solutions are contradictory or cannot be realized, such a combination should not be considered to exist, and is not within the protection scope of the present application.
Fig. 1 schematically shows a flowchart of a method for improving face recognition accuracy according to an embodiment of the present application. As shown in fig. 1, an embodiment of the present application provides a method for improving face recognition accuracy, where the method may include the following steps:
102, acquiring candidate face information according to the face information;
103, respectively determining the probability of punching a card at the current terminal equipment, the probability of punching a card at the current time period and the probability of being correctly identified by the face corresponding to each candidate face information according to a historical database;
and 105, determining the face with the highest face matching accuracy as the target face information.
In the embodiment of the application, the existing attendance machine for the face recognition technology in industrial production usually records a face base picture in advance, extracts face features and establishes a face file. The attendance machine acquires the face picture in real time, compares the face picture with the face file, and takes the comparison result as the face recognition result. However, the accuracy of recognition is low due to the large number of the face base libraries, and meanwhile, the quality of the pictures uploaded to the face base libraries by different departments is different, and the accuracy of face comparison is also affected by the pictures with poor quality. Therefore, the embodiment of the application comprehensively utilizes the existing department organization data, the employee personal behavior data and the face comparison result data to improve the accuracy of face recognition.
In an embodiment of the application, a server is provided, which communicates with a plurality of terminal devices. The terminal equipment can be face recognition attendance machines arranged in a plurality of places. The server may include a face base library and a history database. The human face base library stores human face pictures needing human face recognition; the historical database can be established according to the time information of the historical card punching faces uploaded by the plurality of terminal devices and the terminal device information.
Under the condition that the terminal device detects that a card punching behavior exists, firstly, face information of a current card punching is obtained, the face information is transmitted to the server, and after the server receives the face information sent by the terminal device, candidate face information is determined in a face bottom library according to the obtained face information. The candidate face information is a plurality of pieces of face information with higher similarity to the face information of the current card punching in the face base, wherein the number of the candidate face information can be set according to the actual situation. For example, for each face information, 5 candidate face information having a higher similarity may be set. And the server respectively determines the probability of card punching at the current terminal equipment, the probability of card punching at the current time period and the probability of correctness of the face identification corresponding to each candidate face information according to the historical database, and inputs the probability of card punching at the current terminal equipment, the probability of card punching at the current time period and the probability of correctness of the face identification into the Bayesian model so as to obtain the face matching accuracy of each candidate. The Bayesian model is a prediction by using Bayesian statistics, and not only utilizes model information and data information, but also can fully utilize prior information. The Bayes model is compared with the common regression prediction model through an empirical analysis method, and the result shows that the Bayes model has obvious superiority. After determining the face matching result of each candidate, the server may rank the face matching accuracy of each candidate, determine the candidate face information with the highest face matching accuracy as the target face information, and push the face information to the terminal device.
According to the method and the device, the candidate face information is obtained according to the face information by obtaining the face information of the current card punching. And respectively determining the probability of punching a card at the current terminal equipment, the probability of punching a card at the current time period and the probability of being correctly recognized by the human face corresponding to each candidate human face information according to the historical database. And inputting the probability of punching a card at the current terminal equipment, the probability of punching a card at the current time period and the probability of being correctly recognized by the human face into the Bayesian model to obtain the human face matching accuracy of each candidate. And determining the face with the highest face matching accuracy as the target face information. The embodiment of the application can make full use of the existing attendance card punching behaviors of the staff, matches the staff who probably punch the card to specific equipment, adopts data of a recent period of time for the behaviors of the user in the process of prediction analysis, and performs dynamic statistics on the identified terminal equipment, so that the embodiment of the application has the function of dynamically updating the data prediction result, improves the accuracy of face recognition and reduces the cost.
In this embodiment, the obtaining of the candidate face information according to the face information may include:
and acquiring candidate face information which is most similar to the current card punching face information through a face matching algorithm.
Specifically, since the face information in the face base is more, in order to improve the recognition efficiency, the current card punching face information may be matched in the face base, the face information most similar to the current card punching face information is obtained after matching, and the candidate face information is specified in the most similar face information. The candidate face information can be represented by job numbers or codes. The number of candidate face information can be set according to actual conditions. In the actual process of calculating the face, the closest distance d is generally the most similar, and since normalization processing is performed during calculation, c =1-d can be used as the judgment of the similarity. In one example, the face information of 20 persons is in the face bottom library, and 5 persons with face information most similar to the current face information can be matched as candidate face information.
In this embodiment of the present application, respectively determining, according to the historical database, a probability of punching a card at the current terminal device corresponding to each candidate face information, where the probability of punching a card at the current time period may include:
acquiring equipment point data and time data collected by terminal equipment in a historical database;
determining the probability value of each candidate face information on the current terminal equipment card according to the equipment point data;
and determining the probability value of each candidate face information in the current time slot according to the time data in the historical database.
Specifically, the server can perform predictive analysis of the user card punching behavior by using the existing department organization data, employee personal behavior data and face comparison result data, so as to improve the identification efficiency. The server can firstly obtain the equipment point data and the time data of the terminal equipment, namely the face recognition attendance machine, in the historical database. The probability value of the punching of each candidate face information on the current face recognition attendance machine can be determined by counting the equipment point data in the historical database; the probability value of each candidate face information in the current time slot for card punching can be determined by counting the time data in the historical database.
In this embodiment of the present application, determining, according to the device point data, a probability value of each candidate face information for card punching at the current terminal device may include:
acquiring a code of current terminal equipment;
counting the number of times of punching a card on the current terminal equipment by each candidate face information according to the codes;
and determining the probability value of each candidate face information on the current terminal equipment according to the times of the candidate face information on the current terminal equipment.
Specifically, the equipment logged in by each face in the face file is counted according to the department where the face recognition attendance machine is located and the range of the face to be recognized, then the equipment codes of the attendance machine which need to be transmitted to the server are counted respectively according to the process that the attendance machine accesses the face recognition server, and the probability value of attendance and card punching of the staff at a certain terminal equipment respectively can be obtained according to the login times and the total times of card punching of the staff on each face recognition attendance machine respectively. For example, a certain employee performs a card punch on duty all day for 40 working days in 2 months, wherein the employee effectively performs the card punch on duty on the terminal device a1 for 30 times, effectively performs the card punch on duty on the terminal device a2 for 6 times, effectively performs the card punch on duty on the terminal device a3 for 4 times, and effectively performs the card punch on duty on the terminal device a4 for 0 times (according to the scene, there is no probability of 0 at the lowest point. Namely, the probability values of the card punching of the employee on the terminal devices a1, a2, a3 and a4 are respectively 30/40,6/40,4/40 and 4/40.
In this embodiment of the present application, respectively determining the probability that each candidate face information is corresponding to the face recognition accuracy may include:
acquiring face information input in a face bottom library;
acquiring face information input by terminal equipment;
calculating the probability of correct face recognition corresponding to each candidate face information according to the face information input in the face base and the face information input by the terminal equipment
Specifically, the server first obtains the face information of each candidate pre-stored in the face library, then obtains the face information of the current terminal device, and calculates the probability that each candidate face information is respectively matched with the face information of the current terminal device according to the face information of each candidate pre-stored in the face library and the face information of the current terminal device. For example, there are 20 people in the face base, and in the 20 people base, face information 5 people most similar to the current card face information are matched (the face coding distance is closest, and can be understood as the most similar). In the process of actually calculating the face, the closest distance d is generally the most similar, and since normalization processing is performed during calculation, c =1-d can be used as a similar judgment, so that the matching probabilities of the current card face information and 5 candidate face information are respectively c1, c2, c3, c4 and c5.
In the embodiment of the application, the probability of being correct by face recognition is represented by the face similarity obtained by a face matching algorithm.
Specifically, a face matching algorithm is prestored in the server, and under the condition that face recognition is required, the face similarity of each candidate face information can be respectively obtained through the face matching algorithm, wherein the face similarity is the probability that the candidate face information is correctly recognized by the face.
In the embodiment of the application, the Bayesian model satisfies the formula (1):
wherein, p (x | s, t, c) is the probability of successful matching when certain face information is on certain terminal equipment in a certain time period and under the condition of acquiring the similarity of matched faces; p (s | x) is the probability of the appearance of certain face information in certain face recognition terminal equipment; p (t | x) is the probability of the face information appearing in a certain time period; p (c | x) is the probability that certain face information is correct by face recognition; p (x) is the prior probability of using face recognition service for certain face information; p(s) is the probability value of the card punching of each face information in the terminal equipment; p (t) is a probability value of the possibility of checking attendance and punching cards of each face information in a time period; and p (c) is the face matching similarity.
Specifically, for all people in the enterprise, a person is the same at a certain device, a certain time and possibly matched faces, so that p (x) and p(s) p (t) p (c) are equal, and only the case of p (s | x) p (t | x) p (c | x) p (x) is considered. The further measurement of p (x | s, t, c) is found to correlate only with conditional probabilities about device, time, degree of match. Therefore, the real face to be predicted and recognized can be determined by comparing the sizes of p (x | s, t, c) in different faces. By maximum likelihood estimation: the large occupation advantage of p (x | s, t, c) can be understood as the most similar occupation advantage to the picture provided by the terminal.
In an embodiment of the present application, the method may further include:
and updating the historical database according to the target face information, the corresponding terminal equipment and the time.
In particular, in practical application, for larger-scale companies and enterprises, the existing employee attendance card punching behavior can be fully utilized, and people who may punch the card can be matched to specific equipment, so that the identification accuracy is improved. In the process of prediction analysis, the behavior of the staff adopts data of a recent period of time, and dynamic statistics is carried out on the identified terminal equipment to update a historical database, so that the device for improving the face identification precision in the embodiment of the application has the function of dynamically updating the data prediction result. The statistical frequency can be set according to actual conditions. In one example, the server iterates the device point information and time information in the historical database for the first two months each day, updating the employee's photos in the face base every half year.
In a specific embodiment, the server first performs statistics on the terminal device: employee a often makes a card punch on terminal device a1 (terminal device a1 is a card punch common to employee a). If the time range is set to 2 months, 40 working days are punched on duty, wherein a is effectively punched on terminal a1 for 30 times, 6 times on terminal a2, 4 times on terminal a3, and 0 time on terminal a4 (no probability of 0 is determined according to the scene, the number of punched cards of a certain terminal is set to be the set minimum number of punched cards if the card is not punched on the terminal, specifically, the time range can be set according to the actual service scene, where the card is never punched on terminal a4 of a certain person, but the default value is 4 times). Namely, the probability values of the card punching of the employee A on the terminal devices a1, a2, a3 and a4 are 30/40,6/40,4/40 and 4/40.
The server makes statistics on the time: assuming that the normal working time is 8 am, the card time is divided into 10 minutes before working (which can be divided according to statistics), for example, 7. Assuming that the number of 5 card-punching time periods of the employee on 40 full-duty working days is 5,20,10,4,1, the probability of corresponding time t1-5 in time is 5/40,20/40,10/40,4/40,1/40.
The server calculates the face matching similarity: assuming 20 persons in the face base (actually, there are many persons, which is a relatively large set, and this is an example), in order to improve the calculation efficiency, in the 20 person base, the face matched with the face most similar to the current card is 5 persons, a, B, C, D, and E (the face coding distance is closest, which may also be understood as the most similar). In the actual process of calculating the face, the closest distance d is generally the most similar, and since normalization processing is performed during calculation, c =1-d can be used as the judgment of similarity, and the face similarity degrees c1, c2, c3, c4, and c5 are calculated for 5 persons respectively.
In face recognition, probability values of five persons A, B, C, D and E are calculated respectively. The calculation method is as follows: assuming that a is clicked on the terminal device a1 of 7 min 46 min, the probability value converted according to the bayesian formula is P1=30/40 × 10/40 × c1. The probability of identifying B is then P2= (conditional probability of B on the terminal device) (conditional probability of B in the time period) × C2, and so on C, D, E. The determination that the calculation result is the largest is the result of recognition.
According to the embodiment of the application, the face information of the current card punching is obtained, and the candidate face information is obtained according to the face information. And respectively determining the probability of punching a card at the current terminal equipment, the probability of punching a card at the current time period and the probability of being correctly recognized by the human face corresponding to each candidate human face information according to the historical database. And inputting the probability of punching a card at the current terminal equipment, the probability of punching a card at the current time period and the probability of being correctly recognized by the human face into the Bayesian model to obtain the human face matching accuracy of each candidate. And determining the face with the highest matching accuracy as the target face information. The embodiment of the application can make full use of the existing attendance card punching behaviors of the staff, matches the staff who probably punch the card to specific equipment, adopts data of a recent period of time for the behaviors of the user in the process of prediction analysis, and performs dynamic statistics on the identified terminal equipment, so that the embodiment of the application has the function of dynamically updating the data prediction result, improves the accuracy of face recognition and reduces the cost.
Fig. 2 schematically shows a block diagram of a server according to an embodiment of the present application. As shown in fig. 2, an embodiment of the present application provides a server, which may include:
a memory 210 configured to store instructions; and
the processor 220 is configured to call instructions from the memory 210 and when executing the instructions can implement the above-described method for improving the accuracy of face recognition.
Specifically, in the present embodiment, the processor 220 may be configured to:
acquiring face information of a current card punch;
acquiring candidate face information according to the face information;
respectively determining the probability of punching a card at the current terminal equipment, the probability of punching a card at the current time period and the probability of being correctly recognized by the human face corresponding to each candidate human face information according to a historical database;
inputting the probability of punching a card at the current terminal equipment, the probability of punching a card at the current time period and the probability of being correctly recognized by the human face into a Bayesian model to obtain the human face matching accuracy of each candidate;
and determining the face with the highest matching accuracy as the target face information.
Further, the processor 220 may be further configured to:
and obtaining candidate face information which is most similar to the current card punching face information through a face matching algorithm.
Further, the processor 220 may be further configured to:
acquiring equipment point data and time data collected by terminal equipment in a historical database;
determining the probability value of each candidate face information on the current terminal equipment card according to the equipment point data;
and determining the probability value of each candidate face information in the current time slot according to the time data in the historical database.
Further, the processor 220 may be further configured to:
acquiring a code of current terminal equipment;
counting the number of times of punching a card on the current terminal equipment by each candidate face information according to the codes;
and determining the probability value of each candidate face information on the current terminal equipment according to the times of the candidate face information on the current terminal equipment.
Further, the processor 220 may be further configured to:
acquiring face information input in a face bottom library;
acquiring face information input by terminal equipment;
and calculating the probability of correct face recognition corresponding to each candidate face information according to the face information input in the face base and the face information input by the terminal equipment.
In the embodiment of the application, the probability of being correct by face recognition is represented by the face similarity obtained by a face matching algorithm.
In the embodiment of the application, the Bayesian model satisfies the formula (1):
wherein, p (x | s, t, c) is the probability of successful matching when certain face information is on certain terminal equipment in a certain time period and under the condition of acquiring the similarity of matched faces; p (s | x) is the probability of the appearance of certain face information in certain face recognition terminal equipment; p (t | x) is the probability of the appearance of certain face information in a certain time period; p (c | x) is the probability that certain face information is correct by face recognition; p (x) is the prior probability of using face recognition service for certain face information; p(s) is the probability value of each face information card punching on the terminal equipment; p (t) is a probability value of the possibility of checking attendance and punching cards of each face information in a time period; and p (c) is the face matching similarity.
Further, the processor 220 may be further configured to:
and updating the historical database according to the target face information and the corresponding terminal equipment and time.
By the technical scheme, the face information of the current card punching is obtained, and the candidate face information is obtained according to the face information. And respectively determining the probability of punching a card at the current terminal equipment, the probability of punching a card at the current time period and the probability of being correctly recognized by the human face corresponding to each candidate human face information according to the historical database. And inputting the probability of punching a card at the current terminal equipment, the probability of punching a card at the current time period and the probability of being correctly recognized by the human face into the Bayesian model to obtain the human face matching accuracy of each candidate. And determining the face with the highest face matching accuracy as the target face information. The method and the device can make full use of the existing staff attendance card punching behaviors, match the possible card punching personnel to specific equipment, adopt the data of a recent period of time for the behaviors of the user in the process of predictive analysis, and perform dynamic statistics on the identified terminal equipment, have the function of dynamically updating the data prediction result, improve the accuracy of face recognition and reduce the cost.
Fig. 3 schematically shows a structure diagram of an apparatus for improving accuracy of face recognition according to an embodiment of the present application, and as shown in fig. 3, an apparatus for improving accuracy of face recognition according to an embodiment of the present application may include:
a plurality of terminal devices 310; and
according to the server 320 described above.
Specifically, the terminal device obtains face information of a current card punch, the server 320 obtains candidate face information according to the face information, and then determines probability of the current terminal device card punch, probability of the current time slot card punch and probability of correct face recognition corresponding to each candidate face information according to the historical database. And inputting the probability of checking the card at the current terminal equipment, the probability of checking the card at the current time period and the probability of being correctly recognized by the human face into a Bayesian model to obtain the human face matching accuracy of each candidate, and finally determining the highest human face matching accuracy as the target human face information and pushing the target human face information to the terminal equipment. The method and the device can make full use of the existing staff attendance card punching behaviors, match the possible card punching personnel to specific equipment, adopt the data of a recent period of time for the behaviors of the user in the process of predictive analysis, and perform dynamic statistics on the identified terminal equipment, have the function of dynamically updating the data prediction result, improve the accuracy of face recognition and reduce the cost.
The embodiment of the application also provides a machine-readable storage medium, which stores instructions for causing a machine to execute the above method for controlling the boom.
As will be appreciated by one skilled in the art, embodiments of the present application may be provided as a method, system, or computer program product. Accordingly, the present application may take the form of an entirely hardware embodiment, an entirely software embodiment or an embodiment combining software and hardware aspects. Furthermore, the present application may take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, and the like) having computer-usable program code embodied therein.
The present application is described with reference to flowchart illustrations and/or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the application. It will be understood that each flow and/or block of the flow diagrams and/or block diagrams, and combinations of flows and/or blocks in the flow diagrams and/or block diagrams, can be implemented by computer program instructions. These computer program instructions may be provided to a processor of a general purpose computer, special purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create means for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be stored in a computer-readable memory that can direct a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer-readable memory produce an article of manufacture including instruction means which implement the function specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
In a typical configuration, a computing device includes one or more processors (CPUs), input/output interfaces, network interfaces, and memory.
The memory may include forms of volatile memory in a computer readable medium, random Access Memory (RAM) and/or non-volatile memory, such as Read Only Memory (ROM) or flash memory (flash RAM). The memory is an example of a computer-readable medium.
Computer-readable media, including both non-transitory and non-transitory, removable and non-removable media, may implement information storage by any method or technology. The information may be computer readable instructions, data structures, modules of a program, or other data. Examples of computer storage media include, but are not limited to, phase change memory (PRAM), static Random Access Memory (SRAM), dynamic Random Access Memory (DRAM), other types of Random Access Memory (RAM), read Only Memory (ROM), electrically Erasable Programmable Read Only Memory (EEPROM), flash memory or other memory technology, compact disc read only memory (CD-ROM), digital Versatile Discs (DVD) or other optical storage, magnetic cassettes, magnetic tape, magnetic disk storage or other magnetic storage devices, or any other non-transmission medium that can be used to store information that can be accessed by a computing device. As defined herein, a computer readable medium does not include a transitory computer readable medium such as a modulated data signal and a carrier wave.
It should also be noted that the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, method, article, or apparatus. Without further limitation, an element defined by the phrases "comprising one of 8230; \8230;" 8230; "does not exclude the presence of additional like elements in a process, method, article, or apparatus that comprises the element.
The above are merely examples of the present application and are not intended to limit the present application. Various modifications and changes may occur to those skilled in the art. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application should be included in the scope of the claims of the present application.
Claims (11)
1. A method for improving face recognition accuracy, the method comprising:
acquiring face information of a current card punch;
acquiring candidate face information according to the face information;
respectively determining the probability of punching a card at the current terminal equipment, the probability of punching a card at the current time period and the probability of being correctly identified by the face corresponding to each candidate face information according to a historical database;
inputting the probability of punching the card at the current terminal equipment, the probability of punching the card at the current time period and the probability of being correct for the face identification into a Bayesian model to obtain the face matching accuracy of each candidate;
and determining the face with the highest matching accuracy as the target face information.
2. The method of claim 1, wherein the obtaining candidate face information from the face information comprises:
and obtaining candidate face information which is most similar to the current card punching face information through a face matching algorithm.
3. The method according to claim 1, wherein the determining, according to the historical database, the probability of a card punch at the current terminal device and the probability of a card punch at the current time period corresponding to each candidate face information respectively comprises:
acquiring equipment point data and time data collected by the terminal equipment in the historical database;
determining the probability value of each candidate face information on the current terminal equipment card according to the equipment point data;
and determining the probability value of each candidate face information in the current time slot for card punching according to the time data in the historical database.
4. The method of claim 3, wherein the determining, from the device point data, a probability value of each candidate face information being punched on a current terminal device comprises:
acquiring the code of the current terminal equipment;
counting the number of times of card punching of each candidate face information on the current terminal equipment according to the codes;
and determining the probability value of each candidate face information on the current terminal equipment according to the times of the card punching of each candidate face information on the current terminal equipment.
5. The method of claim 1, wherein the separately determining the probability that each candidate face information is correctly identified by the face comprises:
acquiring face information input in a face bottom library;
acquiring face information input by the terminal equipment;
and calculating the probability of correct face recognition corresponding to each candidate face information according to the face information input in the face base and the face information input by the terminal equipment.
6. The method of claim 5, wherein the probability of being correct for face recognition is represented by a face similarity obtained by a face matching algorithm.
7. The method of claim 1, wherein the bayesian model satisfies formula (1):
wherein, p (x | s, t, c) is the probability of successful matching of certain face information on certain terminal equipment in a certain time period under the condition of acquiring the similarity of matched faces; p (s | x) is the probability of the appearance of the certain face information in the certain face identification terminal equipment; p (t | x) is the probability of the appearance of certain face information in a certain time period; p (c | x) is the probability that certain face information is correct by face recognition; p (x) is a prior probability of using face recognition service for certain face information; p(s) is the probability value of the card punching of each face information in the terminal equipment; p (t) is a probability value of the possibility of checking attendance and punching cards of each face information in a time period; and p (c) is the face matching similarity.
8. The method of claim 1, further comprising:
and updating the historical database according to the target face information, the corresponding terminal equipment and the time.
9. A server, comprising:
a memory configured to store instructions; and
a processor configured to call the instructions from the memory and when executing the instructions to implement the method for improving face recognition accuracy according to any one of claims 1 to 8.
10. An apparatus for improving accuracy of face recognition, comprising:
a plurality of terminal devices; and
the server of claim 9.
11. A machine-readable storage medium having stored thereon instructions for causing a machine to perform the method for improving face recognition accuracy according to any one of claims 1 to 8.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210940098.3A CN115393926A (en) | 2022-08-05 | 2022-08-05 | Method and device for improving face recognition precision and server |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210940098.3A CN115393926A (en) | 2022-08-05 | 2022-08-05 | Method and device for improving face recognition precision and server |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN115393926A true CN115393926A (en) | 2022-11-25 |
Family
ID=84119290
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202210940098.3A Pending CN115393926A (en) | 2022-08-05 | 2022-08-05 | Method and device for improving face recognition precision and server |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN115393926A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116434313A (en) * | 2023-04-28 | 2023-07-14 | 北京声迅电子股份有限公司 | Face recognition method based on multiple face recognition modules |
-
2022
- 2022-08-05 CN CN202210940098.3A patent/CN115393926A/en active Pending
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116434313A (en) * | 2023-04-28 | 2023-07-14 | 北京声迅电子股份有限公司 | Face recognition method based on multiple face recognition modules |
| CN116434313B (en) * | 2023-04-28 | 2023-11-14 | 北京声迅电子股份有限公司 | Face recognition method based on multiple face recognition modules |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| KR102315497B1 (en) | Method and device for building a scoring model and evaluating user credit | |
| CN110431560B (en) | Target person searching method, device, equipment and medium | |
| US20190019197A1 (en) | Determining to dispatch a technician for customer support | |
| CN108694657B (en) | Client identification apparatus, method and computer-readable storage medium | |
| CN110688974B (en) | Identity recognition method and device | |
| CN109214421B (en) | Model training method and device and computer equipment | |
| CN112052733A (en) | Database construction method, face recognition device and electronic equipment | |
| US20230153845A1 (en) | System and method for generating custom data models for predictive forecasting | |
| CN112632609A (en) | Abnormality detection method, abnormality detection device, electronic apparatus, and storage medium | |
| CN112507212A (en) | Intelligent return visit method and device, electronic equipment and readable storage medium | |
| CN112819611A (en) | Fraud identification method, device, electronic equipment and computer-readable storage medium | |
| CN115393926A (en) | Method and device for improving face recognition precision and server | |
| CN113538154A (en) | Risk object identification method and device, storage medium and electronic equipment | |
| US20190050473A1 (en) | Mixture model based time-series clustering of crime data across spatial entities | |
| CN112765463B (en) | Data management method for big data and user requirements and cloud computing server | |
| US20230333720A1 (en) | Generating presentation information associated with one or more objects depicted in image data for display via a graphical user interface | |
| CN112465012A (en) | Machine learning modeling method and device, electronic equipment and readable storage medium | |
| CN116610821A (en) | Knowledge graph-based enterprise risk analysis method, system and storage medium | |
| CN111277465A (en) | Abnormal data message detection method and device and electronic equipment | |
| CN110276648A (en) | Client's identification and based reminding method and device | |
| CN113468886B (en) | Work order processing method and device and computer equipment | |
| CN115409104A (en) | Method, apparatus, device, medium and program product for identifying object type | |
| CN115082041A (en) | User information management method, device, equipment and storage medium | |
| CN109885710B (en) | User image depicting method based on differential evolution algorithm and server | |
| CN113870010A (en) | Data processing method, device and equipment based on machine learning and storage medium |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |