Lightweight unmanned aerial vehicle face recognition method based on integrated learning strategy

Abstract

The invention belongs to the technical field of unmanned aerial vehicle face recognition, and particularly relates to a lightweight unmanned aerial vehicle face recognition method based on an integrated learning strategy. Firstly, an unmanned aerial vehicle face image is collected; thirdly, constructing a lightweight human face detection model, and extracting a low-resolution human face image from the human face image of the unmanned aerial vehicle by using the lightweight human face detection model; constructing a face super-resolution reconstruction model, and extracting a high-resolution face image from the low-resolution face image by using the face super-resolution reconstruction model; constructing a three-dimensional face reconstruction model, reconstructing a face image from the high-resolution face image by using the three-dimensional face reconstruction model, and generating a three-dimensional face reconstruction image; and finally, a lightweight face recognition model DAMobileFace based on a double attention mechanism and a lightweight face recognition model CAMFace based on a CA attention mechanism are constructed. And face recognition is performed on the high-resolution face image and the three-dimensional face reconstruction image by using a DAMobile model and a CAMface model, and the recognition results of the two models on the two images are weighted based on an integrated learning strategy to obtain a final face recognition result. According to the method, the problems of low resolution, complex background, limited computing resources and the like existing in unmanned aerial vehicle face recognition are solved.

Description

Technical Field

[0001] The present invention belongs to the technical field of face recognition, and in particular is a lightweight unmanned aerial vehicle face recognition method based on an integrated learning strategy. Background Art

[0002] With the rapid development of science and technology, drone technology is closely integrated with face recognition technology and is widely used in the fields of intelligent transportation, video surveillance, security tracking and border security. However, drones need to maintain a certain distance from the target during flight and are affected by complex environments, resulting in low quality facial images, complex backgrounds and small face sizes. In addition, drones have limited computing resources, which seriously affects the accuracy of face detection and recognition.

[0003] Current research mainly relies on conventional face recognition models trained directly on drone face datasets. Although a certain degree of accuracy has been achieved on specific datasets, the unique challenges faced by drone face recognition have not been fully considered. First, face images taken by drones usually have low-resolution characteristics, resulting in the loss of facial details and difficulty in extracting effective features; second, drone shooting angles are variable, and facial postures and lighting environments are complex, which increases the difficulty of recognition; third, the drone's own computing resources are limited and it is difficult to support high-complexity model reasoning. Conventional face recognition models, such as Facenet and ArcFace, perform well on high-resolution face images in a stable environment, but cannot achieve satisfactory results on low-resolution drone face images. Therefore, the present invention proposes a lightweight drone face recognition method based on an ensemble learning strategy. Summary of the invention

[0004] In view of the shortcomings of existing face recognition technology, the technical problem that the present invention intends to solve is to propose a lightweight UAV face recognition method based on an ensemble learning strategy.

[0005] The present invention solves the technical problem by adopting the following technical solution:

[0006] A lightweight UAV face recognition method based on an ensemble learning strategy, characterized in that the method comprises the following steps:

[0007] Step 1: Collect drone face images;

[0008] Step 2: Build a lightweight face detection model and use it to extract low-resolution face images from drone face images;

[0009] Step 3: Build a super-resolution face reconstruction model and use it to extract a high-resolution face image from a low-resolution face image.

[0010] Step 4: construct a 3D face reconstruction model, and use the 3D face reconstruction model to reconstruct the face image from the high-resolution face image to generate a 3D face reconstruction image;

[0011] Step 5: Based on the FaceNet network, the backbone network of the Facenet network is replaced with the MobileNet V2 network, and the CBAM module is embedded after the first convolution block and the last bottleneck block of the MobileNet V2 network, and a SENet module is embedded after the deep convolution of each bottleneck block to obtain the lightweight face recognition model DAMobileFace based on the dual attention mechanism;

[0012] Based on the MixNets network, a channel shuffle module is embedded after the element-by-element addition operation of each 3×3, 5×5, 7×7MixConv and 3×3, 5×5, 7×7MixConv-d of the MixNets network, and a CA attention module is embedded after the first and second 3×3, 5×5, 7×7MixConv and the first 3×3, 5×5, 7×7, 9×9MixConv, respectively, to obtain a lightweight face recognition model CAMFace based on the CA attention mechanism;

[0013] The DAMobileface and CAMface models are used to perform face recognition on high-resolution face images and 3D face reconstruction images, respectively. Assume that the accuracy of the DAMobileface model on the high-resolution face image dataset and the 3D face reconstruction image dataset is A super,DAM and A recon,DAM The accuracy of the CAMface model on the high-resolution face image dataset and the 3D face reconstruction image dataset are A super,CAM and A recon,CAM , the accuracy of each model is standardized according to the following formula;

[0014]

[0015] In the formula, max(·) means taking the maximum value, W super,DAM , W recon,DAM are the weights of the DAMobileface model on the high-resolution face image dataset and the 3D face reconstruction image dataset, respectively. super,CAM , W recon,CAM They are the weights of the CAMface model on the high-resolution face image dataset and the 3D face reconstruction image dataset;

[0016] Assume that the recognition results of the DAMobileface model on high-resolution face images and 3D face reconstruction images are Rsuper,DAM and R recon,DAM The recognition results of CAMface model on high-resolution face images and 3D face reconstruction images are R super,CAM and R recon,CAM , then the final face recognition result R is calculated by the following formula;

[0017] R=W super,DAM ·R super,DAM +W recon,DAM ·R recon,DAM +W super,CAM ·R super,CAM +W recon,CAM ·R recon,CAM (6)

[0018] Furthermore, the lightweight face detection model is obtained by replacing each CSP module of the YOLO V5 model backbone network one by one using a lightweight cross-scale feature fusion module; the lightweight cross-scale feature fusion module is divided into a main path and a residual path, the main path includes two convolution modules, multiple RepConv modules and a convolution module connected in sequence, the residual path performs 1×1 convolution on the input features of the module, and the output features of the main path and the residual path are added and then subjected to 1×1 convolution to obtain the output features of the lightweight cross-scale feature fusion module.

[0019] Furthermore, the face super-resolution reconstruction model is obtained by replacing each basic residual block of the generator of the SRGAN network one by one with the residual dense block fused with the attention mechanism; the residual dense block fused with the attention mechanism introduces the attention module CBAM on the basis of the residual dense block RDB, and the attention module CBAM is located after the 1×1 convolution of the residual dense block RDB, and the output features of the attention module CBAM are residually connected with the input features of the dense residual block RDB to obtain the output features of the residual dense block fused with the attention mechanism.

[0020] Furthermore, the three-dimensional face reconstruction model performs three-dimensional face reconstruction based on 3DMM, in which a face instance is represented by a pair of shape vectors and texture vectors, and a PCA algorithm is used to describe the spatial distribution of the shape vectors and texture vectors;

[0021] t=t0+B t δ(1)

[0022] s=s0+B id α+B exp β(2)

[0023] Where t and s are shape vector and texture vector respectively, s0 and t0 are average shape vector and average texture vector respectively, B id , B exp and Bt is the basis matrix of the PCA algorithm, α, β and δ represent the face reconstruction coefficients;

[0024] The face reconstruction coefficient also includes the illumination parameter γ and the posture matrix p, and the face reconstruction coefficient is represented by the vector v = (α, β, δ, γ, p); the ResNet-50 network is used to regress the face reconstruction coefficient vector v to obtain a three-dimensional face reconstructed image.

[0025] Compared with the prior art, the present invention has the following beneficial effects:

[0026] In the face recognition task based on drone images, due to the complex and changeable environment (for example, different flight altitudes, shooting angles, face angles, lighting conditions, etc.), the traditional single recognition method is difficult to maintain high accuracy in complex and changeable scenes. Therefore, the present invention adopts an integrated learning strategy, and significantly improves the accuracy of face recognition from the perspective of drones by weighting the recognition results of multiple face recognition models in different face images. In view of the low resolution and complex background problems of drone images, face super-resolution reconstruction is adopted to effectively improve the clarity of drone face images; three-dimensional face reconstruction technology is used to restore the three-dimensional face structure from a single image, which effectively enhances the face features. The high-resolution face image provides rich two-dimensional face features for face recognition and three-dimensional face reconstruction, and the three-dimensional face reconstruction image provides rich three-dimensional face features for face recognition. Through the synergistic complementarity of two-dimensional face features and three-dimensional face features, the recognition accuracy is further improved. BRIEF DESCRIPTION OF THE DRAWINGS

[0027] Figure 1 It is the overall flow chart of the present invention;

[0028] Figure 2 4 is a structural diagram of a lightweight cross-scale feature fusion module of the present invention;

[0029] Figure 3 A structural diagram of a face super-resolution reconstruction model of the present invention;

[0030] Figure 4 It is a structural diagram of the lightweight face recognition model DAMobileFace based on the dual attention mechanism of the present invention;

[0031] Figure 5 It is a structural diagram of the lightweight face recognition model CAMFace based on the CA attention mechanism of the present invention;

[0032] Figure 6 This is a comparison chart of face recognition accuracy of different methods. DETAILED DESCRIPTION

[0033] The technical solution of the present invention is described in detail below with reference to specific drawings and embodiments, but the protection scope of the present application is not limited thereto.

[0034] The present invention provides a lightweight UAV face recognition method based on an ensemble learning strategy (hereinafter referred to as the method, see Figures 1 to 6 ), including the following steps:

[0035] Step 1: Collect drone face images;

[0036] DJI drones were used to collect videos of people in different states in playgrounds, basketball courts, streets, parking lots, teaching buildings and other places. The drones were hovered at a height of 4 to 8 meters. Each video was about 1 minute long and was processed at a speed of 30FPS. A total of about 6,000 drone face images were obtained. The face target boxes in each image were manually annotated using annotation software, and the annotation files were matched with the faces to obtain a drone face image dataset.

[0037] The drone face images collected from the same video are uniformly divided into training sets or test sets, resulting in a training set of about 4,800 images and a test set of 1,200 images. The face scale of the data set is concentrated below 60 pixels to reflect the characteristics of face images from the drone's perspective.

[0038] Step 2: Improve the YOLOV5 model, build a lightweight face detection model, and train the model. Use the trained lightweight face detection model to extract low-resolution face images from drone face images.

[0039] A lightweight cross-scale feature fusion module is introduced into the YOLO V5 model. The lightweight cross-scale feature fusion module is used to replace each CSP module of the YOLO V5 model backbone network one by one to further extract and enhance the features of each layer and process more abstract semantic information, thereby obtaining a lightweight face detection model. The lightweight cross-scale feature fusion module is divided into a main path and a residual path. The main path includes two convolution modules, multiple RepConv modules and a convolution module connected in sequence. The residual path performs a 1×1 convolution on the input features of the module. The output features of the main path and the residual path are added and then subjected to a 1×1 convolution to obtain the output features of the lightweight cross-scale feature fusion module. The lightweight cross-scale feature fusion module can effectively integrate detailed features and contextual information, and enhance the model's ability to extract features for small-scale faces.

[0040] The NWD (Normalized Wasserstein Distance) loss function is used to calculate the model training loss. NWD uses Wasserstein distance and normalization design to achieve a smoother optimization effect and greater sensitivity to the scale of the target box in target detection tasks. Therefore, the NWD loss function is used instead of the original IoU loss. Compared with the IoU loss, NWD is more sensitive to the relative position and scale difference between the prediction box and the target box, making the optimization process more stable, which further improves the accuracy of the face detection model.

[0041] Step 3: Improve the SRGAN network by introducing the residual dense block with fusion attention mechanism into the SRGAN network, and use the residual dense block with fusion attention mechanism to replace each basic residual block of the generator of the SRGAN network one by one to build a face super-resolution reconstruction model; train the face super-resolution reconstruction model, and use the trained face super-resolution reconstruction model to extract high-resolution face images from low-resolution face images;

[0042] The residual dense block with fusion attention mechanism introduces the attention module CBAM based on the residual dense block RDB. The attention module CBAM is located after the 1×1 convolution of the residual dense block RDB. The output features of the attention module CBAM are residually connected with the input features of the dense residual block RDB to obtain the output features of the residual dense block with fusion attention mechanism.

[0043] The attention module (CBAM) focuses on the key parts of the face by assigning dynamic weights to different areas of the face image to solve problems such as uneven illumination, noise, and resolution differences in drone face images. The attention module CBAM consists of two parts: spatial attention and channel attention. The input features of the dynamic attention module Perform global average pooling and global maximum pooling operations in the spatial dimension to generate channel average pooling features And channel maximum pooling features Average pooling of channel features And channel maximum pooling features The attention weight matrix W is calculated by two shared fully connected layers respectively. avg and W max ;in, W1 and W2 are learnable weight matrices, σ(·) represents the Sigmoid activation function; the two attention weights W avg and W max Weighted with the input feature to obtain the channel attention feature F c ; For channel attention feature F c Perform global average pooling and global maximum pooling operations on the channel dimension to generate spatial average pooling features And spatial maximum pooling features The spatial average pooling feature And spatial maximum pooling features After concatenation, the spatial attention weight matrix W is obtained through convolution and Sigmoid activation function. s ;in, Conv(·) represents the convolution operation, [;] represents the concatenation operation; the channel attention feature F c and the spatial attention weight matrix W s Perform element-by-element multiplication to obtain the output feature F of the attention module out ;in, Represents an element-wise multiplication operation.

[0044] When training the face super-resolution reconstruction model, perceptual loss, edge loss, and adversarial loss are introduced to optimize the generator performance. Perceptual loss enhances structural consistency, edge loss improves contour clarity, and adversarial loss optimizes the generator's ability to generate high-frequency details. The high-resolution image obtained by the face super-resolution reconstruction model has a resolution four times higher than the original low-resolution image, which can better display detailed features and improve image quality.

[0045] Step 4: construct a 3D face reconstruction model, train the 3D face reconstruction model, and use the trained 3D face reconstruction model to reconstruct the face image from the high-resolution face image to generate a 3D face reconstruction image;

[0046] The 3D face reconstruction model is based on 3DMM (3D face deformation statistical model) to perform 3D face reconstruction. The 3D face reconstruction is achieved by regressing the face reconstruction coefficient. In 3DMM, the complete 3D face is accurately described by a pair of shape vectors t and texture vectors s. The PCA algorithm is used to describe the spatial distribution of the shape vectors t and texture vectors s. Then a face instance can be expressed as:

[0047] t=t0+B t δ(1)

[0048] s=s0+B id α+B exp β(2)

[0049] Where s0 and t0 represent the average shape vector and average texture vector respectively; B id , B exp and B t is the basis matrix of the PCA algorithm, representing identity, expression and texture respectively; α, β and δ represent the 3DMM coefficients, which are the coefficients of face reconstruction to be regressed, and different coefficients represent different faces;

[0050] The parameters to be regressed also include the illumination parameter γ and the posture matrix p, and the face reconstruction coefficients to be regressed are represented by the vector v = (α, β, δ, γ, p); the face reconstruction coefficients are regressed using the ResNet-50 network, and the face reconstruction coefficients are obtained to obtain a 3D face reconstructed image;

[0051] Image-level loss function and perception-level loss function are introduced to perform weakly supervised learning training on the 3D face reconstruction model to improve the accuracy and robustness of face reconstruction; the image-level loss function focuses on the color information of each pixel in the face image and sparse landmarks, which helps the model effectively handle face reconstruction tasks under occlusion and extreme posture conditions. The image-level loss function is as follows:

[0052]

[0053] Where, L photo (x) is the image level loss, A i represents the skin attention mask of the i-th image, I i represents the i-th original image, I′ i (x) represents the i-th 3D face reconstructed image, is the obtainable reprojected face area, ||.|| represents the L2 norm;

[0054] However, relying solely on image-level loss functions may cause the model to fall into a local minimum during the reconstruction process, resulting in inaccurate reconstruction results. To overcome this challenge, a perceptual-level loss function is introduced during the training process. By calculating the cosine distance between features, the perceptual-level loss function can effectively optimize the texture details and shape accuracy of the reconstructed face, ensuring that the 3D face reconstruction image is more visually realistic and accurate. The perceptual-level loss function is as follows:

[0055]

[0056] Where, L per (x) is the perceptual level loss, f(I) represents the deep feature vector of the original image I, f(I′(x)) represents the deep feature vector of the reconstructed image I′(v), and <·> represents the vector inner product.

[0057] Step 5: Construct a lightweight face recognition model DAMobileFace based on the dual attention mechanism and a lightweight face recognition model CAMface based on the CA attention mechanism. The two models perform face recognition on high-resolution face images and three-dimensional face reconstruction images respectively, and use an integrated learning strategy to fuse the recognition results of the two models, thereby further improving the recognition accuracy while maintaining a low number of parameters.

[0058] like Figure 4As shown in the figure, based on the FaceNet network, the backbone network of the Facenet network is replaced with the MobileNet V2 network, and the CBAM module and SENet module are introduced into the MobileNet V2 network to improve the feature extraction capability. The CBAM module is embedded after the first convolution block and the last bottleneck block of the MobileNetV2 network, and a SENet module is embedded after the deep convolution of each bottleneck block, thus obtaining the lightweight face recognition model DAMobileFace based on the dual attention mechanism. The CBAM module enables the model to focus more on the significant features of the face area, and the SENet module further improves the network's ability to focus on effective features.

[0059] like Figure 5 As shown in the figure, based on the MixNets network, a channel shuffle module (Channel-shuffle) and a CA attention module are introduced into the MixNets network. A channel shuffle module is embedded after each element-by-element addition operation of 3×3, 5×5, 7×7MixConv and 3×3, 5×5, 7×7MixConv-d, and a CA attention module is embedded after the first and second 3×3, 5×5, 7×7MixConv and the first 3×3, 5×5, 7×7, 9×9MixConv, respectively, to obtain a lightweight face recognition model CAMFace based on the CA attention mechanism. The channel shuffle module enhances the feature expression capability, and the CA attention module further highlights the key features of the face by dynamically adjusting the channel weights, so that the network can focus more on effective information, such as the salient areas and detailed features of the face.

[0060] The DAMobileface and CAMface models were trained on high-resolution face image datasets and 3D face reconstruction image datasets at the same time. The accuracy of the DAMobileface model on the high-resolution face image dataset and 3D face reconstruction image dataset were A and B respectively. super,DAM and A recon,DAM The accuracy of the CAMface model on the high-resolution face image dataset and the 3D face reconstruction image dataset are A super,CAM and A recon,CAM ;

[0061] Each accuracy rate is standardized according to the following formula and mapped to the interval [0,1] to obtain the weight of each model on the high-resolution face image dataset and the 3D face reconstruction image dataset;

[0062]

[0063] In the formula, max(·) means taking the maximum value, W super,DAM , W recon,DAM are the weights of the DAMobileface model on the high-resolution face image dataset and the 3D face reconstruction image dataset, respectively. super,CAM , W recon,CAM They are the weights of the CAMface model on the high-resolution face image dataset and the 3D face reconstruction image dataset;

[0064] Assume that the recognition results of the DAMobileface model on high-resolution face images and 3D face reconstruction images are R super,DAM and R recon,DAM The recognition results of CAMface model on high-resolution face images and 3D face reconstruction images are R super,CAM and R recon,CAM , then the final face recognition result R is calculated by the following formula:

[0065] R=W super,DAM ·R super,DAM +W recon,DAM ·R recon,DAM +W super,CAM ·R super,CAM +W recon,CAM ·R recon,CAM (6)

[0066] The final face recognition result R is the weighted combination of the recognition results of the two lightweight face recognition models for different face images. Figure 6The method of the present invention is compared with the existing method (see "Rostami M, Farajollahi A, Parvin H. Deep learning-based face detection and recognition on drones [J]. Journal of Ambient Intelligence and Humanized Computing, 2024, 15 (1): 373-387.") for face recognition on drone images. The results show that the face recognition accuracy of the method of the present invention is significantly higher than that of the existing method, which improves the face recognition performance under the perspective of drones. The recognition accuracy on the three-dimensional face reconstructed image is higher than the recognition accuracy on the high-resolution face image alone, which proves the effectiveness of adding the three-dimensional face reconstruction step after extracting the high-resolution face image. The method of the present invention combines the two stages of face super-resolution reconstruction and three-dimensional face reconstruction. The face super-resolution reconstruction provides rich two-dimensional face features for the subsequent three-dimensional face reconstruction and integrated learning stages, and the three-dimensional face reconstruction provides rich three-dimensional face features for the subsequent face recognition. Through the synergistic complementarity of two-dimensional face features and three-dimensional face features, the accuracy of drone face recognition is significantly improved.

[0067] Any matters not described in the present invention are applicable to the prior art.

Claims

1. A lightweight UAV face recognition method based on ensemble learning strategy, characterized in that: The method comprises the following steps: Step 1: Collect drone face images; Step 2: Build a lightweight face detection model and use it to extract low-resolution face images from drone face images; Step 3: Build a super-resolution face reconstruction model and use it to extract a high-resolution face image from a low-resolution face image. Step 4: construct a 3D face reconstruction model, and use the 3D face reconstruction model to reconstruct the face image from the high-resolution face image to generate a 3D face reconstruction image; Step 5: Based on the FaceNet network, the backbone network of the Facenet network is replaced with the MobileNetV2 network, and the CBAM module is embedded after the first convolution block and the last bottleneck block of the MobileNetV2 network, and a SENet module is embedded after the deep convolution of each bottleneck block to obtain the lightweight face recognition model DAMobileFace based on the dual attention mechanism; Based on the MixNets network, a channel shuffle module is embedded after the element-by-element addition operation of each 3×3, 5×5, 7×7MixConv and 3×3, 5×5, 7×7MixConv-d of the MixNets network, and a CA attention module is embedded after the first and second 3×3, 5×5, 7×7MixConv and the first 3×3, 5×5, 7×7, 9×9MixConv, respectively, to obtain a lightweight face recognition model CAMFace based on the CA attention mechanism; The DAMobileface and CAMface models are used to perform face recognition on high-resolution face images and 3D face reconstruction images, respectively. Assume that the accuracy of the DAMobileface model on the high-resolution face image dataset and the 3D face reconstruction image dataset is A super,DAM and A recon,DAM The accuracy of the CAMface model on the high-resolution face image dataset and the 3D face reconstruction image dataset are A super,CAM and A recon,CAM , the accuracy of each model is standardized according to the following formula; In the formula, max(·) means taking the maximum value, W super,DAM , W recon,DAM are the weights of the DAMobileface model on the high-resolution face image dataset and the 3D face reconstruction image dataset, respectively. super,CAM , W recon,CAM They are the weights of the CAMface model on the high-resolution face image dataset and the 3D face reconstruction image dataset; Assume that the recognition results of the DAMobileface model on high-resolution face images and 3D face reconstruction images are R super,DAM and R recon,DAM The recognition results of CAMface model on high-resolution face images and 3D face reconstruction images are R super,CAM and R recon,CAM , then the final face recognition result R is calculated by the following formula; R=W super,DAM ·R super,DAM +W recon,DAM ·R recon,DAM +W super,CAM ·R super,CAM +W recon,CAM ·R recon,CAM (6)。 2. The lightweight UAV face recognition method based on ensemble learning strategy according to claim 1 is characterized in that: The lightweight face detection model is obtained by replacing each CSP module of the YOLO V5 model backbone network one by one with a lightweight cross-scale feature fusion module; the lightweight cross-scale feature fusion module is divided into a main path and a residual path, the main path includes two convolution modules, multiple RepConv modules and a convolution module connected in sequence, the residual path performs 1×1 convolution on the input features of the module, and the output features of the main path and the residual path are added and then subjected to 1×1 convolution to obtain the output features of the lightweight cross-scale feature fusion module.

3. The lightweight UAV face recognition method based on ensemble learning strategy according to claim 1 or 2, characterized in that: The face super-resolution reconstruction model is obtained by replacing each basic residual block of the generator of the SRGAN network one by one with the residual dense block fused with the attention mechanism; the residual dense block fused with the attention mechanism introduces an attention module CBAM on the basis of the residual dense block RDB, and the attention module CBAM is located after the 1×1 convolution of the residual dense block RDB, and the output features of the attention module CBAM are residually connected with the input features of the dense residual block RDB to obtain the output features of the residual dense block fused with the attention mechanism.

4. The lightweight UAV face recognition method based on ensemble learning strategy according to claim 3 is characterized in that: The three-dimensional face reconstruction model performs three-dimensional face reconstruction based on 3DMM, in which a face instance is described through a pair of shape vectors and texture vectors, and a PCA algorithm is used to describe the spatial distribution of the shape vectors and texture vectors; t=t0+B t δ(1) s=s0+B id α+B exp β(2) Where t and s are shape vector and texture vector respectively, s0 and t0 are average shape vector and average texture vector respectively, B id , B exp and B t is the basis matrix of the PCA algorithm, α, β and δ represent the face reconstruction coefficients; The face reconstruction coefficient also includes the illumination parameter γ and the posture matrix p, and the face reconstruction coefficient is represented by the vector v = (α, β, δ, γ, p); the ResNet-50 network is used to regress the face reconstruction coefficient vector v to obtain a three-dimensional face reconstructed image.

Images