Face recognition processing method and device, storage medium and processor

Abstract

The application discloses a face recognition processing method and device, a storage medium and a processor. The method comprises the following steps: inputting an image to be recognized into a feature extractor to extract face features, and obtaining face feature information; determining an adversarial regression cost function according to the face feature information; processing the face feature information extracted by the feature extractor through the adversarial regression cost function, and obtaining face target feature information, wherein the posture information has been removed from the face target feature information; and performing face recognition on the image through the face target feature information. The application solves the technical problem that the face recognition effect is poor in the prior art.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of image processing, in particular to a face recognition processing method and device, a storage medium and a processor. BACKGROUND

[0002] With the development of technology, face recognition is applied more and more widely. In the face recognition technology, the following methods are used. 1. 3D conversion method, 3D model construction is complex, and the time consumption is large in the forward prediction process. When the face image is converted to 3D, the high-level semantic expression is not considered, and only the bottom layer texture alignment is considered, so the accuracy of the face recognition result is low. 2. Multi-task learning method. Because the face recognition itself is time-consuming to train, the time cost increases after using multi-task learning. Therefore, the development cost is high, and the face features learned by different task learning need to be mapped to the same metric space, which is difficult in itself, so the performance improvement is limited. 3. Classification-based adversarial learning. Using adversarial learning, simply classifying "front" and "side" against each other, this method simply classifies the pose of the face into "front" and "side", which is a discrete multiple class. But the pose itself is a continuous process, so it affects the calculation performance. This method needs to generate a front face picture, and the generation process of the front face picture is also a time-consuming and laborious process that is difficult to train and control, so the development is high.

[0003] At present, no effective solution has been proposed for the above problems. SUMMARY

[0004] The embodiments of the present application provide a face recognition processing method and device, a storage medium and a processor to at least solve the technical problem that the effect of face recognition in the prior art is not good.

[0005] According to one aspect of the embodiments of the present application, a face recognition processing method is provided, which includes: inputting an image to be recognized into a feature extractor to extract face features, to obtain face feature information; determining an adversarial regression cost function according to the face feature information; processing the face feature information extracted by the feature extractor through the adversarial regression cost function, to obtain face target feature information, wherein the pose information has been removed from the face target feature information; and performing face recognition on the image through the face target feature information.

[0006] Further, determining the adversarial regression cost function according to the face feature information includes: inputting the face feature information into a pose discriminator to extract pose features, to obtain pose information; and determining the adversarial regression cost function through the pose information and the face feature information.

[0007] Further, the determining the adversarial regression cost function based on the pose information and the facial feature information comprises: inputting the pose information and the facial feature information into a pose adversarial network to determine the adversarial regression cost function.

[0008] Further, the inputting the pose information and the facial feature information into the pose adversarial network to determine the adversarial regression cost function comprises: obtaining facial real pose information and facial average pose information of the image; and inputting the pose information, the facial feature information, the facial real pose information and the facial average pose information into the pose adversarial network to determine the adversarial regression cost function.

[0009] Further, the determined adversarial regression cost function is:

[0010]

[0011] wherein x is the image, R' is the facial feature information, D is the pose information output by the pose discriminator, l is the facial real pose information, and is the facial average pose information.

[0012] According to another aspect of the embodiments of the present application, a processing method for face recognition is also provided, which comprises: receiving a service call request sent by a client, wherein the service call request carries an image to be recognized; responding to the service call request, performing face recognition on the image in a server; and outputting a face recognition result.

[0013] Further, the performing face recognition on the image in the server comprises: inputting the image into a feature extractor in the server to perform facial feature extraction, to obtain facial feature information; determining an adversarial regression cost function based on the facial feature information; processing the facial feature information extracted by the feature extractor through the adversarial regression cost function, to obtain facial target feature information, wherein the facial target feature information has removed pose information; and performing face recognition on the image through the facial target feature information.

[0014] According to another aspect of the embodiments of the present application, a processing device for face recognition is also provided, which comprises: a first extraction unit configured to input an image to be recognized into a feature extractor to perform facial feature extraction, to obtain facial feature information; a first determination unit configured to determine an adversarial regression cost function based on the facial feature information; a first processing unit configured to process the facial feature information extracted by the feature extractor through the adversarial regression cost function, to obtain facial target feature information, wherein the facial target feature information has removed pose information; and a first recognition unit configured to perform face recognition on the image through the facial target feature information.

[0015] ​Further, the first determining unit further comprises: a first obtaining subunit, configured to input the face feature information into the pose discriminator to perform pose feature extraction to obtain pose information; and a first determining subunit, configured to determine the adversarial regression cost function by using the pose information and the face feature information.

[0016] Further, the first determining subunit further comprises: a first determining module, configured to input the pose information and the face feature information into the pose adversarial network to determine the adversarial regression cost function.

[0017] Further, the first determining module comprises: a first obtaining submodule, configured to obtain face real pose information and face average pose information of the image; and a first determining submodule, configured to input the pose information, the face feature information, the face real pose information and the face average pose information into the pose adversarial network to determine the adversarial regression cost function.

[0018] Further, the determined adversarial regression cost function is as follows:

[0019]

[0020] wherein, x is the image, R' is the face feature information, D is the pose information output by the pose discriminator, l is the face real pose information, and is the face average pose information.

[0021] According to another aspect of the embodiments of the present application, a storage medium is also provided, which comprises a stored program, wherein the program, when running, controls a device where the storage medium is located to perform the method according to any one of the preceding embodiments.

[0022] According to another aspect of the embodiments of the present application, a processor is also provided, which is used to run a program, wherein the program, when running, performs the method according to any one of the preceding embodiments.

[0023] In the embodiments of the present application, the image to be recognized is input into the feature extractor to perform face feature extraction to obtain face feature information; the adversarial regression cost function is determined according to the face feature information; the face feature information extracted by the feature extractor is processed by using the adversarial regression cost function to obtain face target feature information, wherein the pose information is removed from the face target feature information; and the face recognition is performed on the image by using the face target feature information, thereby solving the technical problem of poor face recognition effect in the prior art. The pose information of the face feature information is removed by using the adversarial regression cost function, thereby avoiding the problem of affecting the accuracy and speed of face recognition due to the presence of the pose information in the face feature information, so that the face recognition is performed on the image by using the face target feature information from which the pose information is removed, thereby improving the accuracy and speed of face recognition, and thereby improving the effect of face recognition. ​BRIEF DESCRIPTION OF DRAWINGS

[0024] The accompanying drawings, which are included to provide a further understanding of the application and are incorporated in and constitute a part of this application, illustrate embodiments of the application and together with the description serve to explain the application. In the drawings:

[0025] Figure 1 is a hardware structure block diagram of a computer terminal according to an embodiment of the application;

[0026] Figure 2 is a flow chart of a face recognition processing method according to an embodiment of the application;

[0027] Figure 3 is a schematic diagram of an optional face recognition processing method according to an embodiment of the application;

[0028] Figure 4 is a schematic diagram of a face recognition processing device according to an embodiment of the application;

[0029] Figure 5 is a schematic diagram of an optional face recognition processing method according to an embodiment of the application; and

[0030] Figure 6 is a hardware structure block diagram of an optional computer terminal according to an embodiment of the application. DETAILED DESCRIPTION

[0031] In order to make the technical personnel of the present application better understand the present application, the following will be combined with the drawings of the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, not all. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative labor should be within the scope of protection of the present application.

[0032] It should be noted that the terms "first", "second", etc. in the specification and claims of the present application and the above-mentioned drawings are used to distinguish similar objects, and do not necessarily describe a specific order or sequence. It should be understood that the data thus used can be interchanged under appropriate circumstances, so that the embodiments of the application described herein can be implemented in an order other than that illustrated or described herein. In addition, the terms "include" and "have" and any variations thereof are intended to cover non-exclusive inclusion, for example, a process, method, system, product or device including a series of steps or units does not necessarily limit to those steps or units clearly listed, but can include other steps or units not clearly listed or inherent to these processes, methods, products or devices.

[0033] Embodiment 1

[0034] According to an embodiment of the present application, a face recognition processing method is provided. It should be noted that the steps shown in the flowchart of the accompanying drawings can be executed in a computer system such as a set of computer executable instructions, and although a logical order is shown in the flowchart, in some cases, the steps shown or described herein can be executed in an order different from that shown herein.

[0035] The face recognition processing method provided by the embodiment of the present application can be executed in a mobile terminal, a computer terminal or similar computing device. Figure 1 A hardware structure block diagram of a computer terminal (or mobile device) for implementing the face recognition processing method is shown. As shown in Figure 1 , the computer terminal 10 (or mobile device 10) can include one or more processors 102 (the processor 102 can include but not limited to a microprocessor MCU or a programmable logic device FPGA processing device), a memory 104 for storing data, and a transmission device for communication function. In addition, it can also include a display, an input / output interface (I / O interface), a universal serial bus (USB) port (which can be included as one of the ports of the I / O interface), a network interface, a power supply and / or a camera. Those skilled in the art can understand that Figure 1 The structure shown is only schematic, which does not limit the structure of the above-mentioned electronic device. For example, the computer terminal 10 can include more or less components than those shown in Figure 1 , or have a different configuration than that shown in Figure 1 .

[0036] It should be noted that the one or more processors 102 and / or other data processing circuits described above can be referred to herein as "data processing circuits" in general. The data processing circuit can be embodied in whole or in part as software, hardware, firmware or any other combination. In addition, the data processing circuit can be a single independent processing module, or any one of the other elements combined into the computer terminal 10 (or mobile device) in whole or in part. As referred to in the embodiments of the present application, the data processing circuit as a kind of processor control (for example, the selection of the variable resistance terminal path connected with the interface).

[0037] The memory 104 can be used to store software programs of application software and modules, such as program instructions / data storage devices corresponding to the face recognition processing method of the embodiments of the present application. The processor 102 executes various functional applications and data processing by running the software programs and modules stored in the memory 104, i.e., implements the face recognition processing method of the above-mentioned application program. The memory 104 can include a high-speed random access memory, and can also include a non-volatile memory, such as one or more magnetic storage devices, flash memories, or other non-volatile solid-state memories. In some examples, the memory 104 can further include a memory remotely arranged with respect to the processor 102, which can be connected to the computer terminal 10 through a network. Examples of the above-mentioned network include, but are not limited to, the Internet, an intranet, a local area network, a mobile communication network, and combinations thereof.

[0038] The transmission device 106 is used to receive or send data via a network. Specific examples of the above-mentioned network can include a wireless network provided by a communication provider of the computer terminal 10. In one example, the transmission device 106 includes a network adapter (Network Interface Controller, NIC), which can be connected to other network devices through a base station so as to be able to communicate with the Internet. In one example, the transmission device 106 can be a radio frequency (Radio Frequency, RF) module, which is used to communicate with the Internet in a wireless manner.

[0039] The display can be, for example, a touch screen type liquid crystal display (LCD), which can enable a user to interact with the user interface of the computer terminal 10 (or mobile device).

[0040] Under the above-mentioned operating environment, the present application provides a face recognition processing method as shown in Figure 2 Figure 2 is a flowchart of the face recognition processing method according to the first embodiment of the present application.

[0041] In step S101, the image to be recognized is input into a feature extractor to extract face features, and face feature information is obtained.

[0042] For example, the face feature information extracted above includes biological feature information of a face in the image to be recognized, such as facial information, iris information, etc., and the face feature information also includes face posture information, such as left face posture information and right face posture information, etc.

[0043] In step S102, an adversarial regression cost function is determined according to the face feature information.

[0044] ​Optionally, in the face recognition processing method provided in Embodiment One of the present application, the determining of the adversarial regression cost function according to the face feature information comprises: inputting the face feature information into a pose discriminator to perform pose feature extraction to obtain pose information; and determining the adversarial regression cost function through the pose information and the face feature information.

[0045] The pose discriminator described above will regress the pose as accurately as possible, and by inputting the face feature information into the pose discriminator to perform pose feature extraction, the real pose information of the face in the image to be recognized can be obtained. Through the real pose information output by the pose discriminator and the face feature information output by the feature extractor, the adversarial regression cost function can be determined, which is an adversarial processing process for the real pose information output by the pose discriminator and the face feature information output by the feature extractor. In the adversarial process, the feature extractor will gradually eliminate the pose information in the face recognition feature, so as to make it more pose-robust.

[0046] In step S103, the face feature information extracted by the feature extractor is processed through the adversarial regression cost function to obtain face target feature information, wherein the pose information has been removed from the face target feature information.

[0047] In step S104, the image is subjected to face recognition through the face target feature information.

[0048] By removing the pose information of the face feature information through the adversarial regression cost function, the problem of affecting the accuracy and speed of face recognition due to the presence of pose information in the face feature information is avoided, so that the image is subjected to face recognition through the face target feature information from which the pose information has been removed, thereby improving the accuracy and speed of face recognition and the technical effect of improving the effect of face recognition.

[0049] Optionally, in the face recognition processing method provided in Embodiment One of the present application, the determining of the adversarial regression cost function through the pose information and the face feature information comprises: inputting the pose information and the face feature information into a pose adversarial network to determine the adversarial regression cost function.

[0050] As shown in Figure 3 pose information and the face feature information into a pose adversarial network to determine the adversarial regression cost function, the pose information of the face feature information output by the feature extractor is removed through the adversarial regression cost function, and the face target feature information from which the pose information has been removed is subjected to face recognition through a recognition classifier and a face recognition loss function, thereby improving the accuracy and speed of face recognition.

[0051] Optionally, in the face recognition processing method provided in Embodiment One of the present application, the pose information and the face feature information are input into the pose adversarial network to determine the adversarial regression cost function, comprising: obtaining face real pose information and face average pose information of the image; inputting the pose information, the face feature information, the face real pose information and the face average pose information into the pose adversarial network to determine the adversarial regression cost function.

[0052] Through the above method, the determined adversarial regression cost function is:

[0053]

[0054] wherein x is the image, R' is the face feature information, D is the pose information output by the pose discriminator, l is the face real pose information, and is the face average pose information.

[0055] As can be seen from the above algorithm, in the adversarial process, the feature extractor gradually eliminates the pose information in the face recognition feature, so that it is more robust to pose.

[0056] In summary, the technical solution in the present application does not need to generate a frontal face. It saves computing resources and development costs during training, and reduces prediction time during prediction. In addition, since multi-task training is not used, only one pose discriminator is added compared with single-task naive face recognition training method, and the computational amount of the pose discriminator is only a small proportion (for example, about 5%) of that of the feature extractor, so it greatly saves the computational amount compared with the multi-task method. At the same time, the pose adversarial network used in the pose adversarial process is a continuous function, not a discrete classification network, which is more in line with the nature of the pose being continuous information, thereby having stronger performance, thereby improving the accuracy and speed of face recognition.

[0057] It should be noted that, for each of the above method embodiments, in order to simply describe, each is described as a combination of a series of actions, but those skilled in the art should know that the present application is not limited by the order of the described actions, because according to the present application, certain steps can be performed in other order or simultaneously. Secondly, those skilled in the art should know that the embodiments described in the specification all belong to preferred embodiments, and the actions and modules involved are not necessarily necessary for the present application.

[0058] ​Those skilled in the art can clearly understand the method according to the above-mentioned embodiments can be realized by means of software and necessary general hardware platform, of course, also can be through hardware, but in many cases the former is a better implementation. Based on such understanding, the technical solutions of the present application essentially or say the part of the prior art to make contributions can be in the form of a software product, the computer software product stored in a storage medium (such as ROM / RAM, magnetic disk, optical disc), including a number of instructions to make a terminal device (may be a mobile phone, computer, server, or network equipment, etc.) to execute the method of various embodiments of the present application.

[0059] Embodiment 2

[0060] According to the embodiments of the present application, a device for implementing the above-mentioned face recognition processing method is also provided, as shown in Figure 4 The device comprises a first extraction unit 201, a first determination unit 202, a first processing unit 203, and a first recognition unit 204.

[0061] Specifically, the first extraction unit 201 is configured to input an image to be recognized into a feature extractor to extract face features and obtain face feature information.

[0062] The first determination unit 202 is configured to determine an adversarial regression cost function according to the face feature information.

[0063] The first processing unit 203 is configured to process the face feature information extracted by the feature extractor through the adversarial regression cost function to obtain face target feature information, wherein the pose information has been removed from the face target feature information.

[0064] The first recognition unit 204 is configured to perform face recognition on the image through the face target feature information.

[0065] The face recognition processing device provided in the second embodiment of the present application solves the technical problem of poor face recognition effect in the prior art by inputting the image to be recognized into the feature extractor by the first extraction unit 201 to extract the face feature information, determining the adversarial regression cost function according to the face feature information by the first determination unit 202, processing the face feature information extracted by the feature extractor by the first processing unit 203 through the adversarial regression cost function to obtain the face target feature information, wherein the pose information has been removed from the face target feature information, and performing face recognition on the image through the face target feature information by the first recognition unit 204. The pose information of the face feature information is removed through the adversarial regression cost function, which avoids the problem of affecting the accuracy and speed of face recognition due to the presence of the pose information in the face feature information, so that the accuracy and speed of face recognition are improved by removing the pose information from the face target feature information to perform face recognition on the image, thereby improving the technical effect of the effect of face recognition.

[0066] Optionally, in the face recognition processing device provided in the second embodiment of the present application, the first determination unit 202 further includes: a first acquisition subunit, configured to input the face feature information into the pose discriminator to extract the pose feature information to obtain the pose information; and a first determination subunit, configured to determine the adversarial regression cost function through the pose information and the face feature information.

[0067] Optionally, in the face recognition processing device provided in the second embodiment of the present application, the first determination subunit further includes: a first determination module, configured to input the pose information and the face feature information into the pose adversarial network to determine the adversarial regression cost function.

[0068] Optionally, in the face recognition processing device provided in the second embodiment of the present application, the first determination module includes: a first acquisition sub-module, configured to acquire the face real pose information and the face average pose information of the image; and a first determination sub-module, configured to input the pose information, the face feature information, the face real pose information and the face average pose information into the pose adversarial network to determine the adversarial regression cost function.

[0069] Optionally, in the face recognition processing device provided in the second embodiment of the present application, the determined adversarial regression cost function is:

[0070]

[0071] wherein x is the image, R' is the face feature information, D is the pose information output by the pose discriminator, l is the face real pose information, and is the face average pose information.

[0072] ​It should be noted that the first extraction unit 201, the first determination unit 202, the first processing unit 203, and the first identification unit 204 correspond to steps S101-S104 in Embodiment 1, and the four units have the same instances and application scenarios as the corresponding steps, but are not limited to the content disclosed in Embodiment 1. It should be noted that the above modules can be run in the computer terminal 10 provided in Embodiment 1 as part of the device.

[0073] Embodiment 3

[0074] In the above operating environment, the present application provides a face recognition processing method as shown in Figure 5 Figure 5 is a flowchart of the face recognition processing method according to Embodiment 3 of the present application.

[0075] Step S501, receiving a service call request sent by a client, wherein the service call request carries an image to be identified.

[0076] Step S502, responding to the service call request, performing face recognition on the image in the server.

[0077] Step S503, outputting the face recognition result.

[0078] By calling the service in the server to perform face recognition on the image, the accuracy and speed of face recognition are improved, thereby improving the effect of face recognition.

[0079] Optionally, in the face recognition processing method provided in Embodiment 3 of the present application, performing face recognition on the image in the server includes: inputting the image into a feature extractor in the server to perform face feature extraction, obtaining face feature information; determining an adversarial regression cost function according to the face feature information; processing the face feature information extracted by the feature extractor through the adversarial regression cost function, obtaining face target feature information, wherein the pose information has been removed from the face target feature information; performing face recognition on the image through the face target feature information.

[0080] For example, the face feature information extracted above includes biological feature information of a face in the image to be identified, such as facial information, iris information, etc., and the face feature information also includes face pose information, such as left face pose information and right face pose information, etc.

[0081] By removing the pose information of the face feature information through the adversarial regression cost function, the problem of affecting the accuracy and speed of face recognition due to the presence of pose information in the face feature information is avoided, thereby performing face recognition on the image through the face target feature information from which the pose information has been removed, improving the accuracy and speed of face recognition, and thereby improving the effect of face recognition.​

[0082] Optionally, in the face recognition processing method provided in Embodiment Three of the present application, the determining of the adversarial regression cost function according to the face feature information comprises: inputting the face feature information into the pose discriminator to perform pose feature extraction to obtain pose information; and determining the adversarial regression cost function through the pose information and the face feature information.

[0083] The pose discriminator can regress the pose as accurately as possible, and the real pose information of the face in the image to be recognized can be obtained by inputting the face feature information into the pose discriminator to perform pose feature extraction. The real pose information output by the pose discriminator and the face feature information output by the feature extractor can be used to determine the adversarial regression cost function, which is an adversarial processing process of the real pose information output by the pose discriminator and the face feature information output by the feature extractor. In the adversarial process, the feature extractor gradually eliminates the pose information in the face recognition feature, so that it is more robust to the pose.

[0084] Optionally, in the face recognition processing method provided in Embodiment Three of the present application, the determining of the adversarial regression cost function through the pose information and the face feature information comprises: inputting the pose information and the face feature information into a pose adversarial network to determine the adversarial regression cost function.

[0085] The pose information and the face feature information are input into the pose adversarial network to determine the adversarial regression cost function, and the pose information of the face feature information output by the feature extractor is removed through the adversarial regression cost function. The face target feature information without the pose information is subjected to face recognition through the recognition classifier and the face recognition loss function, so as to improve the accuracy and speed of face recognition.

[0086] Optionally, in the face recognition processing method provided in Embodiment Three of the present application, the inputting of the pose information and the face feature information into the pose adversarial network to determine the adversarial regression cost function comprises: obtaining face real pose information and face average pose information of the image; and inputting the pose information, the face feature information, the face real pose information and the face average pose information into the pose adversarial network to determine the adversarial regression cost function.

[0087] Through the above method, the determined adversarial regression cost function is:

[0088]

[0089] wherein, x is the image, R' is the face feature information, D is the pose information output by the pose discriminator, l is the face real pose information, and is the face average pose information.

[0090] It can be seen from the above algorithm that in the adversarial process, the feature extractor gradually eliminates the pose information in the face recognition feature, so that it is more pose-robust.

[0091] It should be noted that, for the foregoing method embodiments, in order to simply describe, they are all expressed as a series of action combinations, but those skilled in the art should know that the present application is not limited by the action sequence described, because according to the present application, certain steps can be performed in other sequences or simultaneously. Secondly, those skilled in the art should know that the embodiments described in the specification all belong to preferred embodiments, and the actions and modules involved are not necessarily necessary for the present application.

[0092] Through the description of the above embodiments, those skilled in the art can clearly understand that the method according to the above embodiments can be realized by means of software and the necessary general hardware platform, of course, it can also be realized by hardware, but in many cases, the former is a better embodiment. Based on such understanding, the technical solutions of the present application can be embodied in the form of a software product, which is stored in a storage medium (such as ROM / RAM, magnetic disk, optical disk) and includes a plurality of instructions for making a terminal device (which can be a mobile phone, computer, server, or network device, etc.) execute the method of each embodiment of the present application.

[0093] Embodiment 4

[0094] The embodiments of the present application can provide a computer terminal, which can be any computer terminal device in a computer terminal group. Alternatively, in the present embodiment, the above computer terminal can also be replaced by a terminal device such as a mobile terminal.

[0095] Alternatively, in the present embodiment, the above computer terminal can be located in at least one network device of a plurality of network devices of a computer network.

[0096] In the present embodiment, the computer terminal can execute program codes of the following steps in the face recognition processing method of the application program: inputting an image to be recognized into a feature extractor to extract face features, obtaining face feature information; determining an adversarial regression cost function according to the face feature information; processing the face feature information extracted by the feature extractor through the adversarial regression cost function, obtaining face target feature information, wherein the pose information has been removed from the face target feature information; performing face recognition on the image through the face target feature information.

[0097] When executed on a data processing device, it is also suitable to execute an initialization program with the following steps: inputting facial feature information into a pose discriminator to extract pose features and obtain pose information; and determining an adversarial regression cost function based on the pose information and facial feature information.

[0098] When executed on a data processing device, it is also suitable to execute an initialization procedure with the following steps: inputting pose information and facial feature information into the pose adversarial network to determine the adversarial regression cost function.

[0099] When executed on a data processing device, it is also suitable to execute an initialization procedure with the following steps: acquiring the true face pose information and the average face pose information of the image; inputting the pose information, face feature information, true face pose information and average face pose information into the pose adversarial network to determine the adversarial regression cost function.

[0100] When executed on a data processing device, it is also suitable to execute a program that initializes with the following steps: The determined adversarial regression cost function is:

[0101]

[0102] Where x is the image, R' is the facial feature information, D is the pose information output by the pose discriminator, and l is the true facial pose information. It is the average pose information of the human face.

[0103] Optionally, Figure 6 This is a structural block diagram of a computer terminal according to an embodiment of the present invention. Figure 6 As shown, the computer terminal may include: one or more ( Figure 6 (Only one is shown in the image) processor and memory.

[0104] The memory can be used to store software programs and modules, such as the program instructions / modules corresponding to the face recognition processing method and apparatus in this embodiment of the invention. The processor executes various functional applications and data processing by running the software programs and modules stored in the memory, thereby realizing the face recognition processing method described above. The memory may include high-speed random access memory, and may also include non-volatile memory, such as one or more magnetic storage devices, flash memory, or other non-volatile solid-state memory. In some instances, the memory may further include memory remotely located relative to the processor, and these remote memories can be connected to the terminal via a network. Examples of such networks include, but are not limited to, the Internet, corporate intranets, local area networks, mobile communication networks, and combinations thereof.

[0105] The processor can call information and application programs stored in the memory through the transmission device to perform the following steps: inputting an image to be identified into a feature extractor to extract face features to obtain face feature information; determining an adversarial regression cost function according to the face feature information; processing the face feature information extracted by the feature extractor through the adversarial regression cost function to obtain face target feature information, wherein the pose information has been removed from the face target feature information; and performing face recognition on the image through the face target feature information.

[0106] The processor also implements the following steps when executing the program: inputting the face feature information into a pose discriminator to extract pose features to obtain pose information; and determining an adversarial regression cost function through the pose information and the face feature information.

[0107] The processor also implements the following steps when executing the program: inputting the pose information and the face feature information into a pose adversarial network to determine the adversarial regression cost function.

[0108] The processor also implements the following steps when executing the program: obtaining face real pose information and face average pose information of the image; and inputting the pose information, the face feature information, the face real pose information and the face average pose information into the pose adversarial network to determine the adversarial regression cost function.

[0109] The processor also implements the following steps when executing the program: determining that the adversarial regression cost function is:

[0110]

[0111] Wherein, x is the image, R' is the face feature information, D is the pose information output by the pose discriminator, l is the face real pose information, and is the face average pose information.

[0112] The embodiment of the present application provides a face recognition processing method. By inputting an image to be identified into a feature extractor to extract face features to obtain face feature information; determining an adversarial regression cost function according to the face feature information; processing the face feature information extracted by the feature extractor through the adversarial regression cost function to obtain face target feature information, wherein the pose information has been removed from the face target feature information; and performing face recognition on the image through the face target feature information, the technical problem of poor face recognition effect in the prior art is solved. By removing the pose information of the face feature information through the adversarial regression cost function, the problem of affecting the accuracy and speed of face recognition due to the presence of pose information in the face feature information is avoided, so that face recognition is performed on the image through the face target feature information with the pose information removed, the accuracy and speed of face recognition are improved, and the technical effect of improving the effect of face recognition is achieved. ​

[0113] Those skilled in the art can understand that, Figure 6 The structure shown is only schematic, and the computer terminal can also be a smart phone (such as an Android phone, an iOS phone, etc.), a tablet computer, a palm computer, a Mobile Internet Device (MID), a PAD, or the like. Figure 6 It does not limit the structure of the electronic device described above. For example, the computer terminal 10 can further include more or fewer components (such as a network interface, a display device, etc.) than those shown in the figure, or have a different configuration from that shown in the figure. Figure 6 The structure shown is only schematic, and the computer terminal can also be a smart phone (such as an Android phone, an iOS phone, etc.), a tablet computer, a palm computer, a Mobile Internet Device (MID), a PAD, or the like. Figure 6 It does not limit the structure of the electronic device described above. For example, the computer terminal 10 can further include more or fewer components (such as a network interface, a display device, etc.) than those shown in the figure, or have a different configuration from that shown in the figure.

[0114] Those skilled in the art can understand that all or part of the steps in the face recognition processing method of the above-mentioned embodiments can be completed by instructing the hardware related to the terminal device by a program, which can be stored in a computer readable storage medium, which can include a flash disk, a Read-Only Memory (ROM), a Random Access Memory (RAM), a magnetic disk or an optical disk, etc.

[0115] Embodiment 5

[0116] The fourth embodiment of the present application also provides a storage medium. Optionally, in the present embodiment, the storage medium can be used to save the program code executed by the face recognition processing method provided in the first embodiment.

[0117] Optionally, in the present embodiment, the storage medium can be located in any one of the computer terminals in the computer terminal group in the computer network, or in any one of the mobile terminals in the mobile terminal group.

[0118] Optionally, in the present embodiment, the storage medium is configured to store program code for performing the following steps: inputting an image to be recognized into a feature extractor to extract face features, to obtain face feature information; determining an adversarial regression cost function according to the face feature information; processing the face feature information extracted by the feature extractor through the adversarial regression cost function, to obtain face target feature information, wherein the pose information has been removed from the face target feature information; and performing face recognition on the image through the face target feature information.

[0119] The storage medium is also configured to store program code for performing the following steps: determining an adversarial regression cost function according to the face feature information, including: inputting the face feature information into a pose discriminator to extract pose features, to obtain pose information; and determining the adversarial regression cost function through the pose information and the face feature information.

[0120] The storage medium is further configured to store program code for performing the following steps: determining the adversarial regression cost function by the pose information and the face feature information comprises: inputting the pose information and the face feature information into the pose adversarial network to determine the adversarial regression cost function.

[0121] The storage medium is further configured to store program code for performing the following steps: inputting the pose information and the face feature information into the pose adversarial network to determine the adversarial regression cost function comprises: obtaining face real pose information and face average pose information of the image; inputting the pose information, the face feature information, the face real pose information and the face average pose information into the pose adversarial network to determine the adversarial regression cost function.

[0122] The storage medium is further configured to store program code for performing the following steps: the determined adversarial regression cost function is:

[0123]

[0124] Wherein, x is the image, R' is the face feature information, D is the pose information output by the pose discriminator, l is the face real pose information, and is the face average pose information.

[0125] The above-mentioned embodiment numbers of the present application are only for description, and do not represent the advantages and disadvantages of the embodiments.

[0126] In the above-mentioned embodiments of the present application, the description of each embodiment has its own emphasis, and the parts not described in detail in a certain embodiment can be referred to the related description of other embodiments.

[0127] In several embodiments provided in the present application, it should be understood that the disclosed technical contents can be implemented by other ways. Among them, the above-mentioned device embodiments are only schematic, for example, the division of units is only a logical function division, and actual implementation can have another division way, for example, a plurality of units or components can be combined or integrated into another system, or some features can be ignored or not executed. In addition, the coupling or direct coupling or communication connection between the displayed or discussed each other can be through some interface, indirect coupling or communication connection between units or modules, which can be electrical or other forms.

[0128] The units described as separate components can or can not be physically separated, and the components shown as units can or can not be physical units, that is, they can be located in one place, or can be distributed on multiple network units. According to actual needs, part or all of the units can be selected to achieve the purpose of the present embodiment scheme.

[0129] ​In addition, each function unit in each embodiment of the present application can be integrated in one processing unit, or each unit can be physically present separately, or two or more units can be integrated in one unit. The integrated unit can be realized in the form of hardware or in the form of a software function unit.

[0130] If the integrated unit is realized in the form of a software function unit and sold or used as an independent product, it can be stored in a computer readable storage medium. Based on such understanding, the technical solutions of the present application, essentially or the part that contributes to the prior art, or all or part of the technical solutions can be embodied in the form of a software product. The computer software product is stored in a storage medium, and includes several instructions to enable 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 in each embodiment of the present application. The foregoing storage medium includes: a U disk, a read-only memory (ROM, Read-Only Memory), a random access memory (RAM, Random Access Memory), a mobile hard disk, a magnetic disk or an optical disk, and various media that can store program codes.

[0131] The above is only the preferred embodiment of the present application, and it should be pointed out that, for those skilled in the art, without departing from the principles of the present application, a number of improvements and refinements can be made, and these improvements and refinements should be considered as the protection scope of the present application.

Claims

1. A processing method of face recognition, characterized in that, The method comprises the following steps: inputting an image to be identified into a feature extractor to extract facial features, to obtain facial feature information; determining an adversarial regression cost function according to the facial feature information, wherein the adversarial regression cost function is constructed by a first expected value and a second expected value, the first expected value is used to represent an expected value of a deviation between pose information contained in the facial feature information and real pose information of a face in the image, and the second expected value is used to represent an expected value of a deviation between the pose information and average pose information of the face in the image; processing the facial feature information by using the feature extractor through the adversarial regression cost function, to obtain facial target feature information, wherein the pose information has been removed from the facial target feature information; performing face recognition on the image through the facial target feature information.

2. The treatment method according to claim 1, characterized in that, The method comprises the following steps: inputting the facial feature information into a pose discriminator to extract pose features, to obtain the pose information; determining the adversarial regression cost function through the pose information and the facial feature information.

3. The treatment method according to claim 2, characterized in that, The method comprises the following steps: inputting the pose information and the facial feature information into a pose adversarial network to determine the adversarial regression cost function.

4. The treatment method according to claim 3, characterized in that, The method comprises the following steps: obtaining the real pose information and the average pose information of the face in the image according to the pose information; inputting the pose information, the facial feature information, the real pose information and the average pose information of the face into the pose adversarial network to determine the adversarial regression cost function.

5. A processing method of face recognition, characterized by, The method comprises the following steps: receiving a service call request sent by a client, wherein the service call request carries an image to be identified; performing face recognition on the image in a server in response to the service call request; outputting a face recognition result; The method comprises the following steps: inputting the image into a feature extractor in the server to extract facial features, to obtain facial feature information; determining an adversarial regression cost function according to the facial feature information, wherein the adversarial regression cost function is constructed by a first expected value and a second expected value, the first expected value is used to represent an expected value of a deviation between pose information contained in the facial feature information and real pose information of a face in the image, and the second expected value is used to represent an expected value of a deviation between the pose information and average pose information of the face in the image; processing the facial feature information by using the feature extractor through the adversarial regression cost function, to obtain facial target feature information, wherein the pose information has been removed from the facial target feature information; performing face recognition on the image through the facial target feature information.

6. A processing device for face recognition, characterized in that, The method comprises the following steps: a first extraction unit is configured to input an image to be identified into a feature extractor to extract facial features, to obtain facial feature information; The first determining unit is configured to determine an adversarial regression cost function according to the face feature information, wherein the adversarial regression cost function is constructed by a first expected value and a second expected value, the first expected value is used to represent an expected value of a deviation between pose information contained in the face feature information and real pose information of a face in the image, and the second expected value is used to represent an expected value of a deviation between the pose information and average pose information of the face in the image; The first processing unit is configured to process the face feature information by using the feature extractor according to the adversarial regression cost function, to obtain face target feature information, wherein the pose information is removed from the face target feature information; The first identifying unit is configured to perform face identification on the image by using the face target feature information.

7. The processing device of claim 6, wherein, The first determining unit further includes: The first obtaining sub-unit is configured to input the face feature information into a pose discriminator to perform pose feature extraction, to obtain pose information; The first determining sub-unit is configured to determine the adversarial regression cost function by using the pose information and the face feature information.

8. The processing device of claim 7, wherein, The first determining sub-unit further includes: The first determining module is configured to input the pose information and the face feature information into a pose adversarial network, to determine the adversarial regression cost function.

9. The processing device of claim 8, wherein, The first determining module includes: The first obtaining sub-module is configured to obtain real pose information of a face in the image and average pose information of the face according to the pose information; The first determining sub-module is configured to input the pose information, the face feature information, the real pose information of the face, and the average pose information of the face into the pose adversarial network, to determine the adversarial regression cost function.

10. A storage medium, characterized by The storage medium includes a stored program, wherein the program controls a device where the storage medium is located to perform the method in any one of claims 1 to 5 when the program is running.

11. A processor, comprising: The processor is configured to run a program, wherein the program performs the method in any one of claims 1 to 5 when the program is running.

Images