Method, apparatus and electronic device for face recognition

Abstract

The application relates to the technical field of face recognition, and discloses a face recognition method, which comprises the following steps: obtaining a face picture to be recognized; inputting the face picture into a key point detection model and obtaining a plurality of face region pictures output by the key point detection model; determining face feature data according to the plurality of face region pictures; and determining a face recognition result according to the face feature data. In this way, the face picture can be input into the key point detection model to obtain pictures of different face regions, the face feature data can be determined in combination with the pictures of the plurality of face regions, face recognition can be performed in combination with the face feature data, the accuracy of face recognition is effectively improved, and the face recognition method has higher recognition accuracy. The application further discloses a device for face recognition and an electronic equipment.

Description

Technical Field

[0001] This application relates to the field of facial recognition technology, and for example to a method, apparatus and electronic device for facial recognition. Background Technology

[0002] Currently, with the continuous improvement of people's living standards, electronic devices with facial recognition capabilities have emerged. Facial recognition, in this context, is a biometric technology that identifies individuals based on their facial features. As an important biological characteristic of the human body, the face possesses uniqueness and is difficult to replicate. Furthermore, facial images are relatively easy to acquire, thus facial recognition plays a crucial role in many fields.

[0003] Currently, facial recognition has become an essential means of identity authentication in daily life. Most facial recognition algorithms achieve high recognition rates, but their accuracy is not high when the face is obscured, such as when a user is wearing a mask or glasses. To address this issue, users typically need to remove the obstruction before facial recognition, which is cumbersome and, during periods of heightened pandemic activity, hinders pandemic control efforts. Summary of the Invention

[0004] To provide a basic understanding of some aspects of the disclosed embodiments, a brief summary is given below. This summary is not intended as a general commentary, nor is it intended to identify key / important components or describe the scope of protection of these embodiments, but rather as a prelude to the detailed description that follows.

[0005] This disclosure provides a method, apparatus, and electronic device for face recognition, thereby providing a more accurate face recognition method.

[0006] In some embodiments, the face recognition method includes: obtaining a face image to be recognized; inputting the face image into a key point detection model and obtaining multiple face region images output by the key point detection model; determining face feature data based on the multiple face region images; and determining a face recognition result based on the face feature data.

[0007] In some embodiments, the face recognition apparatus includes a processor and a memory storing program instructions, the processor being configured to execute the aforementioned method for face recognition when the program instructions are executed.

[0008] In some embodiments, the electronic device includes a means for facial recognition.

[0009] The method, apparatus, and electronic device for face recognition provided in this disclosure can achieve the following technical effects: By inputting an obtained face image into a key point detection model, multiple face region images output by the key point detection model are obtained. Facial feature data is then determined using these multiple face region images, and the face recognition result is determined by combining the determined facial feature data. In this way, by inputting a face image into a key point detection model to obtain images of different face regions, and combining these multiple face region images to determine more accurate facial feature data, and then performing face recognition using this facial feature data, the accuracy of face recognition is effectively improved, and a face recognition method with higher recognition accuracy is provided.

[0010] The above general description and the description below are exemplary and illustrative only and are not intended to limit this application. Attached Figure Description

[0011] One or more embodiments are illustrated by way of example with reference to the accompanying drawings. These illustrations and drawings do not constitute a limitation on the embodiments. Elements having the same reference numerals in the drawings are shown as similar elements. The drawings are not to be scaled. And wherein:

[0012] Figure 1 This is a schematic diagram of a method for face recognition provided in an embodiment of this disclosure;

[0013] Figure 2 This is a schematic diagram of a method for determining facial feature data provided in an embodiment of this disclosure;

[0014] Figure 3 This is a schematic diagram of another method for determining facial feature data provided in this disclosure embodiment;

[0015] Figure 4 This is a schematic diagram of a method for correcting facial feature data provided in an embodiment of this disclosure;

[0016] Figure 5 This is a schematic diagram of a method for determining face recognition results provided in an embodiment of this disclosure;

[0017] Figure 6 This is a schematic diagram of a face recognition device provided in an embodiment of this disclosure. Detailed Implementation

[0018] To provide a more detailed understanding of the features and technical content of the embodiments of this disclosure, the implementation of the embodiments of this disclosure will be described in detail below with reference to the accompanying drawings. The accompanying drawings are for illustrative purposes only and are not intended to limit the embodiments of this disclosure. In the following technical description, for ease of explanation, several details are used to provide a full understanding of the disclosed embodiments. However, one or more embodiments may still be implemented without these details. In other cases, well-known structures and devices may be simplified in their depiction to simplify the drawings.

[0019] The terms "first," "second," etc., used in the specification, claims, and accompanying drawings of this disclosure are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate for the embodiments of this disclosure described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover non-exclusive inclusion.

[0020] Unless otherwise stated, the term "multiple" means two or more.

[0021] In this embodiment of the disclosure, the character " / " indicates that the objects before and after it are in an "or" relationship. For example, A / B means: A or B.

[0022] The term "and / or" describes an association between objects, indicating that three relationships can exist. For example, A and / or B means: A or B, or A and B.

[0023] The term "correspondence" can refer to an association or binding relationship. The correspondence between A and B means that there is an association or binding relationship between A and B.

[0024] In this embodiment, the electronic device can communicate with smart home appliances via the internet, or directly via Bluetooth, Wi-Fi, or other methods. Here, smart home appliances refer to home appliances that incorporate microprocessors, sensor technology, and network communication technology, possessing characteristics of intelligent control, intelligent sensing, and intelligent applications. The operation of smart home appliances often relies on the application and processing of modern technologies such as the Internet of Things, the internet, and electronic chips. In this embodiment, the electronic device can be a mobile device, a smart home appliance, a computer, or other device with facial recognition capabilities. Mobile devices may include, for example, mobile phones, smart home devices, wearable devices, smart mobile devices, virtual reality devices, or any combination thereof. Wearable devices include, for example, smartwatches, smart bracelets, pedometers, etc.

[0025] Figure 1 This is a schematic diagram of a method for face recognition provided in an embodiment of this disclosure; combined with Figure 1 As shown, this disclosure provides a method for face recognition, including:

[0026] S11, the electronic device obtains a face image to be identified.

[0027] S12, the electronic device inputs the face image into the key point detection model and obtains multiple face region images output by the key point detection model.

[0028] S13, the electronic device determines facial feature data based on multiple facial region images.

[0029] S14, the electronic device determines the face recognition result based on the face feature data.

[0030] In this solution, an electronic device can acquire an image of a face to be identified via its associated image acquisition device. Here, the image acquisition device can be a camera. In one example, if the image acquired by the image acquisition device also includes other body parts of the user, the image can be input into a pre-stored face detection model in the electronic device to process the image. After the electronic device obtains the face location coordinates output by the face detection model, it uses these coordinates to extract the face image from the acquired image. In this way, the image to be identified acquired by the image acquisition device can be processed so that the electronic device can obtain the face image to be identified.

[0031] Furthermore, after obtaining the face image, it can be input into a keypoint detection model. Here, the keypoint detection model is pre-trained and capable of removing facial regions that do not contain facial keypoints from the face image, and cropping out facial regions that do contain facial keypoints. In this embodiment, the input to the keypoint detection model is the face image to be identified, and the output of the keypoint detection model is an image of multiple facial regions. As an example, keypoints can be marked in the face image before training the keypoint detection model. For example, 68 keypoints can be marked in the face image, and these 68 keypoints can be labeled as multiple facial regions. These multiple facial regions can include the eye region, the bridge of the nose region, the mouth region, the upper contour region, the lower contour region, etc. Here, the eye region can include all keypoints at the eyes and eyebrows in the image. In this way, after keypoints are labeled in the face image, the pre-built keypoint detection model can be trained to obtain the keypoint detection model used in this embodiment, thereby outputting multiple facial region images through the keypoint detection model. Specifically, in this scheme, multiple facial region images output by the keypoint detection model can be used, including images of the eyes, bridge of the nose, mouth, upper contour, and lower contour. This allows for face segmentation using the keypoint detection model and provides an accurate data foundation for determining facial feature data.

[0032] Furthermore, multiple facial region images can be combined to determine facial feature data. In this embodiment, after acquiring multiple facial region images, each image can be fed into its corresponding feature extraction model to obtain target feature data for each region. By combining multiple target feature data, facial feature data can be determined, thereby enabling the determination of facial recognition results.

[0033] The face recognition method provided in this disclosure involves inputting an obtained face image into a keypoint detection model to obtain multiple face region images output by the keypoint detection model. Facial feature data is then determined using these multiple face region images, and the face recognition result is determined by combining the determined facial feature data. This method effectively improves the accuracy of face recognition by inputting a face image into a keypoint detection model to obtain images of different face regions, combining these multiple face region images to determine more accurate facial feature data, and then using this facial feature data for face recognition. This provides a face recognition method with higher recognition accuracy.

[0034] Figure 2 This is a schematic diagram of a method for determining facial feature data provided in an embodiment of this disclosure; combined with Figure 2As shown, optionally, in step S13, the electronic device determines facial feature data based on multiple facial region images, including:

[0035] S21, the electronic device obtains the feature extraction models corresponding to each of the multiple facial region images.

[0036] S22, the electronic device inputs multiple facial region images into their respective feature extraction models and obtains target feature data output by multiple feature extraction models.

[0037] S23, the electronic device determines facial feature data based on multiple target feature data.

[0038] In this scheme, the electronic device can obtain feature extraction models corresponding to multiple facial region images. For example, the feature extraction models may include eyebrow models corresponding to the eye region, nose bridge models corresponding to the nose bridge region, mouth models corresponding to the mouth region, upper contour models corresponding to the upper contour region, and lower contour models corresponding to the lower contour region. Specifically, the feature extraction models corresponding to different facial regions can be obtained as follows: Keypoint detection can be performed on a publicly available face dataset to obtain segmented images of multiple regions. Here, the segmented images of multiple regions are images at the same scale to ensure the stability of the extracted features. The segmented images of multiple regions are then fed into a CNN (Convolutional Neural Network) for feature extraction, and after feature extraction, facial features are determined by pre-set weights. The pre-set weights can be:eye:nose bridge:mouth:upper contour:lower contour = 0.7:0.1:0.1:0.05:0.05. That is, facial features = 0.7 * eye features + 0.1 * nose bridge features + 0.1 * mouth features + 0.05 * upper contour features + 0.05 * lower contour features. In this way, by strengthening the eye weights during model training, the features of the eyes can be highlighted. After obtaining facial features, the model can be converged through loss function calculation, and after the model converges, feature extraction models corresponding to different facial regions can be extracted from the network.

[0039] Furthermore, after obtaining feature extraction models corresponding to multiple facial regions, the images of the facial regions to be identified can be input into their respective feature extraction models. For example, the image of the eyes to be identified can be input into the eye feature extraction model, and the image of the bridge of the nose to be identified can be input into the bridge of the nose feature extraction model. In this way, by inputting different facial region images into different feature extraction models, the target feature data output by each model can be obtained. Then, by combining the target feature data output by different feature extraction models, facial feature data can be determined. Here, the target feature data includes: eye features, bridge of the nose features, mouth features, upper contour features, and lower contour features. In this way, by inputting multiple facial region images into their respective feature extraction models, target feature data for each facial region can be obtained, providing a more accurate data foundation for determining facial feature data.

[0040] Figure 3 This is a schematic diagram of another method for determining facial feature data provided in this disclosure embodiment; combined with Figure 3 As shown, in step S23, the electronic device determines facial feature data based on multiple target feature data, including:

[0041] S31, the electronic device determines the feature weights corresponding to each of the multiple target feature data.

[0042] S32, the electronic device calculates the product of each target feature data and its corresponding feature weight, and determines the sum of multiple products as face feature data.

[0043] As an example, multiple target feature data and their corresponding feature weights can be pre-stored in an electronic device based on the reference levels of different facial regions. Understandably, the higher the reference level, the greater the influence of the corresponding facial region features, and the sum of the feature weights corresponding to the multiple target feature data is 1. For example, to emphasize the eye features of a face, the feature weight corresponding to the eye features can be increased. In this embodiment, the feature weights of the target features can be pre-set as eye:nose:mouth:upper contour:lower contour = 0.7:0.1:0.1:0.05:0.05. This approach allows for a focus on referencing eye data during the determination of facial feature data. Furthermore, facial feature data can be determined by combining the feature weights corresponding to the multiple determined target feature data. In one example, the product of each target feature data and its corresponding feature weight can be calculated separately, and the sum of these products can be used to determine the facial feature data. That is, facial feature = 0.7 * eye feature + 0.1 * nose bridge feature + 0.1 * mouth feature + 0.05 * upper contour feature + 0.05 * lower contour feature. This scheme can accurately calculate facial feature data by combining the feature weights of different facial regions.

[0044] Optionally, in step S31, the electronic device determines the feature weights corresponding to each of the multiple target feature data, including:

[0045] When the facial region corresponding to the target feature data is occluded, the electronic device determines the feature weight corresponding to the target feature data as the first weight.

[0046] When there is no occlusion in the facial area corresponding to the target feature data, the electronic device determines the feature weight corresponding to the target feature data as the second weight.

[0047] The first weight is less than the second weight.

[0048] In this scheme, the smaller the feature weight, the less impact the feature has on the accuracy of face recognition. Conversely, the larger the feature weight, the greater the impact on the accuracy of face recognition. This allows for setting smaller feature weights when the face region is occluded and larger feature weights when the face region is unoccluded. Therefore, the feature weight corresponding to the target feature data of the face region can be reduced when the face region is occluded. In one case, if the user is wearing a mask, it can be determined that the mouth region and / or lower contour region are occluded, and the feature weight of the mouth region and / or lower contour region can be reduced. As a preferred option, the first weight can be set to zero. In another case, if the user is wearing glasses, it can be determined that the eyebrow and eye region are occluded, and the feature weight of the eyebrow and eye region can be reduced. This weakens the feature influence of occluded face regions, thereby improving the accuracy of face recognition to some extent when the user's face is occluded. Furthermore, when the face region corresponding to the target feature data is unoccluded, the electronic device can also determine the feature weight corresponding to the target feature data as a second weight. Here, the second weight is greater than the first weight. The second weight can be a pre-defined weight, for example, eyes:nose:mouth:upper contour:lower contour = 0.7:0.1:0.1:0.05:0.05. In this way, the weights of different facial regions can be allocated by combining the pre-defined weight relationship to achieve more accurate face recognition.

[0049] Figure 4 This is a schematic diagram of a method for correcting facial feature data provided in an embodiment of this disclosure; combined with Figure 4 As shown in the embodiments of this disclosure, a method for correcting facial feature data is also provided, including:

[0050] S41, when the face region corresponding to the target feature data is occluded, and the face region is the eye region, the electronic device processes the eye region image among multiple face region images to remove the occluders in the eye region image.

[0051] S42, the electronic device inputs the processed eye region image into the feature extraction model corresponding to the eye region, and obtains new eye feature data output by the model;

[0052] S43, the electronic device corrects facial feature data based on the new eye feature data.

[0053] In this solution, the electronic device can process the eye region image when the facial region corresponding to the target feature data is occluded, and the facial region is specifically the eye region. Specifically, if the user is wearing glasses in the eye region image, then the eye region is determined to be occluded. Further, the eye region image can be input into a preset glasses removal model for processing. Thus, the eye region image output by the preset glasses removal model can be identified as the processed eye region image. In this way, occlusion processing is effectively performed on images with occluded eye regions, allowing for the acquisition of new eye feature data by combining the processed eye region image. Specifically, the obtained processed eye region image can be input into an eye feature extraction model, and the new eye feature data output by the model can be obtained. This new eye feature data is then used to correct the facial feature data. Here, the eye feature data used in the facial feature data calculation process can be replaced with the new eye feature data. This eliminates the need to adjust the feature weights of the eye features and achieves more accurate facial feature data.

[0054] Figure 5 This is a schematic diagram of a method for determining face recognition results provided in an embodiment of this disclosure; combined with Figure 5 As shown, optionally, in step S14, the electronic device determines the face recognition result based on the face feature data, including:

[0055] S51, the electronic device obtains facial database data.

[0056] S52, the electronic device determines the face recognition result based on the comparison result between the face feature data and the face database data.

[0057] In this embodiment, facial recognition database data can be pre-stored in the electronic device or its associated server. This database includes facial feature data of different users and their respective authentication information. Specifically, after acquiring facial feature data, it can be compared with the pre-stored database data in the electronic device to obtain the cosine similarity between the two data sets. This cosine similarity is then used as the comparison result, and the facial recognition result is determined based on this result. In this way, a more accurate facial recognition result can be determined by comparing the cosine similarity between the data sets.

[0058] Optionally, S52, the electronic device determines the face recognition result based on the comparison result between the face feature data and the face database data, including:

[0059] If the comparison result is greater than the first threshold, the electronic device determines that the face recognition result is successful.

[0060] If the comparison result is greater than the second threshold but less than the first threshold, the electronic device determines that the face recognition result is a recognition failure.

[0061] In this embodiment, a first threshold and a second threshold can be preset based on the accuracy requirements of face recognition. As an example, the first threshold can be 0.6 and the second threshold can be 0.5. This way, if the determined comparison result is greater than 0.6, the face recognition result is determined to be successful. If the determined comparison result is greater than 0.5 and less than 0.6, the face recognition result is determined to be unsuccessful. In this way, by combining the preset first and second thresholds to determine the recognition result, the accuracy of the data can be further guaranteed.

[0062] Optionally, if the face recognition result is determined to be a recognition failure, the electronic device receives the identity authentication information sent by the target user.

[0063] If the identity authentication information indicates that the identity recognition is accurate, the electronic device determines that the facial recognition result is successful.

[0064] In this embodiment, the target user is the user to be facially recognized, and the authentication information is information that can identify the user. For example, the authentication information can be the target user's mobile phone number, verification code, name, etc. Specifically, if the facial recognition result is determined to be a failure, the electronic device can receive the authentication information sent by the target user to match the electronic device with the pre-stored authentication information. If the match is successful, the facial recognition result is adjusted to a successful recognition. In this way, the recognition dimensions can be further expanded, and the recognition success rate can be improved.

[0065] This disclosure provides an apparatus for face recognition, including a first obtaining module, a second obtaining module, a first determining module, and a second determining module. The first obtaining module is configured to obtain a face image to be recognized; the second obtaining module is configured to input the face image into a key point detection model and obtain multiple face region images output by the key point detection model; the first determining module is configured to determine face feature data based on the multiple face region images; and the second determining module is configured to determine a face recognition result based on the face feature data.

[0066] The face recognition apparatus provided in this disclosure inputs an acquired face image into a keypoint detection model to obtain multiple face region images output by the keypoint detection model. These multiple face region images are then used to determine face feature data, which is then combined to determine the face recognition result. This approach allows for the input of a face image into a keypoint detection model to obtain images of different face regions. Combining these multiple face region images to determine more accurate face feature data, and then using this face feature data for face recognition, effectively improves the accuracy of face recognition and provides a face recognition method with higher recognition precision.

[0067] Figure 6 This is a schematic diagram of a face recognition device provided in an embodiment of this disclosure; combined with Figure 6 As shown, this disclosure provides an apparatus for face recognition, including a processor 100 and a memory 101. Optionally, the apparatus may further include a communication interface 102 and a bus 103. The processor 100, communication interface 102, and memory 101 can communicate with each other via the bus 103. The communication interface 102 can be used for information transmission. The processor 100 can call logical instructions in the memory 101 to execute the face recognition method of the above embodiment.

[0068] Furthermore, the logic instructions in the aforementioned memory 101 can be implemented as software functional units and, when sold or used as independent products, can be stored in a computer-readable storage medium.

[0069] The memory 101, as a computer-readable storage medium, can be used to store software programs and computer-executable programs, such as program instructions / modules corresponding to the methods in the embodiments of this disclosure. The processor 100 executes functional applications and data processing by running the program instructions / modules stored in the memory 101, that is, it implements the face recognition method in the above embodiments.

[0070] The memory 101 may include a program storage area and a data storage area. The program storage area may store the operating system and applications required for at least one function; the data storage area may store data created based on the use of the terminal device. Furthermore, the memory 101 may include high-speed random access memory and may also include non-volatile memory.

[0071] This disclosure provides an electronic device that includes the above-described device for face recognition.

[0072] This disclosure provides a computer-readable storage medium storing computer-executable instructions configured to perform the above-described method for face recognition.

[0073] This disclosure provides a computer program product, which includes a computer program stored on a computer-readable storage medium. The computer program includes program instructions that, when executed by a computer, cause the computer to perform the above-described method for face recognition.

[0074] The aforementioned computer-readable storage medium may be a transient computer-readable storage medium or a non-transitory computer-readable storage medium.

[0075] The technical solutions of this disclosure can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes one or more instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the method described in this disclosure. The aforementioned storage medium can be a non-transitory storage medium, including: a USB flash drive, a portable hard drive, a read-only memory (ROM), a random access memory (RAM), a magnetic disk, or an optical disk, and other media capable of storing program code; it can also be a transient storage medium.

[0076] The foregoing description and accompanying drawings fully illustrate embodiments of this disclosure to enable those skilled in the art to practice them. Other embodiments may include structural, logical, electrical, procedural, and other changes. The embodiments represent only possible variations. Individual components and functions are optional unless explicitly required, and the order of operation may vary. Parts and features of some embodiments may be included in or replace parts and features of other embodiments. Moreover, the terminology used in this application is for describing embodiments only and is not intended to limit the claims. As used in the description of embodiments and claims, the singular forms “a,” “an,” and “the” are intended to equally include the plural forms unless the context clearly indicates otherwise. Similarly, the term “and / or” as used in this application means including one or more of the associated listed items and all possible combinations thereof. Additionally, when used in this application, the term "comprise" and its variations "comprises" and / or "comprising" refer to the presence of stated features, integrals, steps, operations, elements, and / or components, but do not exclude the presence or addition of one or more other features, integrals, steps, operations, elements, components, and / or groups thereof. Without further limitations, an element defined by the phrase "comprises a..." does not exclude the presence of other identical elements in the process, method, or apparatus that includes said element. In this document, each embodiment may focus on the differences from other embodiments, and similar or identical parts between embodiments can be referred to mutually. For methods, products, etc., disclosed in the embodiments, if they correspond to the method section disclosed in the embodiments, the relevant parts can be referred to the description of the method section.

[0077] Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of the embodiments of this disclosure. Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working processes of the systems, devices, and units described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here.

[0078] The methods and products (including but not limited to devices and equipment) disclosed in the embodiments herein can be implemented in other ways. For example, the device embodiments described above are merely illustrative. For instance, the division of units may be merely a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. In addition, the coupling or direct coupling or communication connection between the shown or discussed units may be through some interfaces, and the indirect coupling or communication connection between devices or units may be electrical, mechanical, or other forms. The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units, that is, they may be located in one place or distributed across multiple network units. Some or all of the units may be selected to implement this embodiment according to actual needs. Furthermore, the functional units in the embodiments of this disclosure may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit.

[0079] The flowcharts and block diagrams in the accompanying drawings illustrate the architecture, functionality, and operation of possible implementations of systems, methods, and computer program products according to embodiments of this disclosure. In this regard, each block in a flowchart or block diagram may represent a module, segment, or portion of code containing one or more executable instructions for implementing a specified logical function. In some alternative implementations, the functions marked in the blocks may occur in a different order than that shown in the drawings. For example, two consecutive blocks may actually be executed substantially in parallel, and they may sometimes be executed in reverse order, depending on the functions involved. In the descriptions corresponding to the flowcharts and block diagrams in the accompanying drawings, the operations or steps corresponding to different blocks may also occur in a different order than disclosed in the description, and sometimes there is no specific order between different operations or steps. For example, two consecutive operations or steps may actually be executed substantially in parallel, and they may sometimes be executed in reverse order, depending on the functions involved. Each block in a block diagram and / or flowchart, and combinations of blocks in a block diagram and / or flowchart, can be implemented using a dedicated hardware-based system that performs the specified function or action, or using a combination of dedicated hardware and computer instructions.

Claims

1. A method for face recognition, characterized in that, include: Obtain the face image to be identified; The face image is input into a keypoint detection model, and multiple face region images are obtained from the keypoint detection model. The keypoint detection model is a model that removes face regions that do not contain facial keypoints from the face image and crops out the face regions that contain facial keypoints. The multiple face region images include eye region images, nose bridge region images, mouth region images, upper contour region images, and lower contour region images. The multiple facial region images are images at the same scale; Based on multiple facial region images, facial feature data is determined, including: obtaining feature extraction models corresponding to each of the multiple facial region images; inputting the multiple facial region images into their respective feature extraction models and obtaining target feature data output by the multiple feature extraction models; determining the feature weights corresponding to each of the multiple target feature data; the feature weights are pre-set based on the reference degree of each facial region, with the reference degree of the eye region being higher than that of the bridge of the nose region, mouth region, upper contour region, and lower contour region; calculating the product of each target feature data and its corresponding feature weight, and determining the sum of multiple products as the facial feature data; The facial recognition result is determined based on the facial feature data.

2. The method according to claim 1, characterized in that, Determining the feature weights corresponding to each of the multiple target feature data includes: If the facial region corresponding to the target feature data is occluded, the feature weight corresponding to the target feature data is determined as the first weight. If the facial region corresponding to the target feature data is not obscured, the feature weight corresponding to the target feature data is determined as the second weight. Wherein, the first weight is less than the second weight.

3. The method according to claim 2, characterized in that, The method further includes: If the face region corresponding to the target feature data is occluded, and the face region is the eye region, the eye region image among the multiple face region images is processed to remove the occluders in the eye region image. The processed image of the eye region is input into the feature extraction model corresponding to the eye region, and new eye feature data output by the model is obtained. The facial feature data is corrected based on the new eye feature data.

4. The method according to claim 1, characterized in that, Determining the face recognition result based on the face feature data includes: Obtain facial recognition database data; The face recognition result is determined based on the comparison between the facial feature data and the facial database data.

5. The method according to claim 4, characterized in that, The step of determining the face recognition result based on the comparison result between the face feature data and the face database includes: If the comparison result is greater than the first threshold, the face recognition result is determined to be successful. If the comparison result is greater than the second threshold and less than the first threshold, the face recognition result is determined to be a recognition failure.

6. The method according to claim 5, characterized in that, Also includes: If the face recognition result is determined to be a recognition failure, the system receives the identity authentication information sent by the target user. If the identity authentication information indicates that the identity recognition is accurate, the face recognition result is determined to be successful.

7. A device for face recognition, comprising a processor and a memory storing program instructions, characterized in that, The processor is configured to execute the method for face recognition as described in any one of claims 1 to 6 when running the program instructions.

8. An electronic device, characterized in that, Includes the device for face recognition as described in claim 7.

Images