CN112446270B - Person re-identification network training method, person re-identification method and device - Google Patents

Abstract

The present application provides a training method for a pedestrian re-identification network, a pedestrian re-identification method and a device. It relates to the field of artificial intelligence, and specifically to the field of computer vision. The method includes: obtaining M training images and the annotation data of the M training images; initializing the network parameters of the pedestrian re-identification network to obtain the initial values of the network parameters of the pedestrian re-identification network; inputting a batch of training images from the M training images into the pedestrian re-identification network for feature extraction, obtaining a feature vector of each training image in the batch of training images, and then determining a loss function based on the feature vectors of the batch of training images, and obtaining a pedestrian re-identification network that meets preset requirements based on the function value of the loss function. The present application can train a pedestrian re-identification network with better performance under the condition of annotated data from a single image capture device.

Description

Person re-identification network training method, person re-identification method and device

Technical Field

The present application relates to the field of computer vision, and more specifically, to a training method for a pedestrian re-identification network, a pedestrian re-identification method and a device.

Background technique

Computer vision is an integral part of various intelligent/autonomous systems in various application fields, such as manufacturing, inspection, document analysis, medical diagnosis, and military. It is a discipline about how to use cameras/image capture devices and computers to obtain the data and information of the objects we need. Figuratively speaking, it is to install eyes (cameras/image capture devices) and brains (algorithms) on computers to replace human eyes to identify, track, and measure targets, so that computers can perceive the environment. Because perception can be seen as extracting information from sensory signals, computer vision can also be seen as a science that studies how to make artificial systems "perceive" from images or multidimensional data. In general, computer vision uses various imaging systems to replace visual organs to obtain input information, and then computers replace the brain to complete the processing and interpretation of these input information. The ultimate research goal of computer vision is to enable computers to observe and understand the world through vision like humans, and have the ability to adapt to the environment autonomously.

The field of surveillance often involves the problem of pedestrian re-identification. Person re-identification (ReID) can also be called pedestrian re-identification. Pedestrian re-identification is a technology that uses computer vision technology to determine whether there is a specific pedestrian in an image or video sequence.

Traditional solutions generally use training data and annotated data across image capture devices to train the pedestrian re-identification network so that the pedestrian re-identification network can distinguish between images of different pedestrians and then identify the pedestrians. However, the training data in the traditional solution includes images of the same pedestrian taken by different image capture devices. Such images taken by different image capture devices need to be manually annotated so that the images of the same pedestrian taken by different image capture devices can be associated (that is, the pedestrians are associated across image capture devices). However, in many scenarios, it is very difficult to associate pedestrians across image capture devices, especially when the number of people and the number of image capture devices increase, the difficulty of cross-image capture device association also increases significantly. Data annotation has high economic costs and consumes a lot of time.

Summary of the invention

The present application provides a training method, a pedestrian re-identification method and a device for a pedestrian re-identification network, so as to train a pedestrian re-identification network with better performance under the condition of annotated data by a single image shooting device.

In a first aspect, a method for training a person re-identification network is provided, the method comprising:

Step 1: Get training data;

The training data in step 1 includes M training images and labeled data of the M training images, where M is an integer greater than 1;

Step 2: Initialize the network parameters of the pedestrian re-identification network to obtain the initial values of the network parameters of the pedestrian re-identification network;

Repeat steps 3 to 5 below until the pedestrian re-identification network meets the preset requirements;

Step 3: Input a batch of training images from the M training images into the person re-identification network for feature extraction to obtain a feature vector for each training image in the batch of training images;

Step 4: Determine the function value of the loss function based on the feature vectors of a batch of training images;

Step 5: Update the network parameters of the person re-identification network according to the function value of the loss function.

In the above step 1, among the M training images of the training data, each training image includes a pedestrian, and the annotation data of each training image includes the bounding box where the pedestrian in each training image is located and the pedestrian identification information. Different pedestrians correspond to different pedestrian identification information. Among the M training images, the training images with the same pedestrian identification information come from the same image capture device. The M training images can be all the training images used when training the pedestrian re-identification network. In the specific training process, a batch of training images from the M training images can be selected each time and input into the pedestrian re-identification network for processing.

The above-mentioned image capturing device may specifically be a device such as a video camera or a still camera that can capture images of pedestrians.

The pedestrian identification information in the above step 1 can also be called pedestrian identity identification information, which is a kind of information used to indicate the identity of the pedestrian. Each pedestrian can correspond to unique pedestrian identification information. There are many ways to represent the pedestrian identification information, as long as it can indicate the identity information of the pedestrian. For example, the pedestrian identification information can specifically be the pedestrian identity (identity, ID), that is, a unique ID can be assigned to each pedestrian.

In the above step 2, the network parameters of the pedestrian re-identification network can be randomly set to obtain the initial values of the network parameters of the pedestrian re-identification network.

In the above step 3, the above batch of training images may include N anchor images, wherein the N anchor images are any N training images in the above batch of training images, and each anchor image in the N anchor images corresponds to a most difficult positive sample image, a first most difficult negative sample image and a second most difficult negative sample image.

The most difficult positive sample image, the first most difficult negative sample image and the second most difficult negative sample image corresponding to each anchor point image are described below.

The most difficult positive sample image corresponding to each anchor image: the training image in the above batch of training images that has the same pedestrian identification information as each anchor image and has the farthest distance from the feature vector of each anchor image;

The first most difficult negative sample image corresponding to each anchor image: a training image in the above batch of training images that comes from the same image capture device as each anchor image, has different pedestrian identification information from each anchor image, and has the closest distance to the feature vector of each anchor image;

The second most difficult negative sample image corresponding to each anchor image: a training image in the above batch of training images that comes from a different image capture device than each anchor image, has different pedestrian identification information from each anchor image, and has the closest distance to the feature vector of each anchor image.

In the above step 4, the function value of the loss function is obtained by averaging the function values of the N first loss functions. The function value of each of the N first loss functions is calculated based on the first difference and the second difference corresponding to each of the N anchor images.

The above N is a positive integer, and the above N is less than M. When N=1, there is only one function value of the first loss function, and at this time, the function value of the first loss function can be directly used as the function value of the loss function in step 4.

Optionally, the function value of each of the first loss functions is the sum of the first difference and the second difference corresponding to each anchor point image.

Optionally, the function value of each of the above-mentioned first loss functions is the sum of the first difference, the second difference and other constant terms corresponding to each anchor point image.

The meanings of the first difference and the second difference and the distances forming the first difference and the second difference are explained below.

The first difference value corresponding to each anchor image: the difference between the most difficult positive sample distance corresponding to each anchor image and the second most difficult negative sample distance corresponding to each anchor image;

The second difference value corresponding to each anchor image: the difference between the second most difficult negative sample distance corresponding to each anchor image and the first most difficult negative sample distance corresponding to each anchor image;

The most difficult positive sample distance corresponding to each anchor image: the distance between the feature vector of the most difficult positive sample image corresponding to each anchor image and the feature vector of each anchor image;

The second most difficult negative sample distance corresponding to each anchor image: the distance between the feature vector of the second most difficult negative sample image corresponding to each anchor image and the feature vector of each anchor image;

The first hardest negative sample distance corresponding to each anchor image: the distance between the feature vector of the first hardest negative sample image corresponding to each anchor image and the feature vector of each anchor image.

In addition, in the present application, several training images coming from the same image capturing device means that these several training images are captured by the same image capturing device.

In the present application, the most difficult negative sample images from different image capturing devices and the same image capturing device are taken into account in the process of constructing the loss function, and the first difference and the second difference are reduced as much as possible during the training process, so as to eliminate the interference of the image capturing device's own information on the image information as much as possible, so that the trained pedestrian re-identification network can more accurately extract features from the image.

Specifically, during the training process of the pedestrian re-identification network, the network parameters of the pedestrian re-identification network are optimized to make the first difference and the second difference as small as possible, thereby making the difference between the most difficult positive sample distance and the second most difficult negative sample distance and the difference between the second most difficult negative sample distance and the first most difficult negative sample distance as small as possible, so that the pedestrian re-identification network can distinguish the features of the most difficult positive image and the second most difficult negative sample image, as well as the features of the second most difficult negative sample image and the first most difficult negative sample image as much as possible, so that the trained pedestrian re-identification network can better and more accurately extract features from the image.

In conjunction with the first aspect, in certain implementations of the first aspect, the pedestrian re-identification network meets the preset requirements, including: when at least one of the following conditions (1) to (3) is met, the pedestrian re-identification network meets the preset requirements:

(1) The number of training times of the person re-identification network is greater than or equal to the preset number;

(2) The function value of the loss function is less than or equal to the preset threshold;

(3) The recognition performance of the pedestrian re-identification network meets the preset requirements.

The above-mentioned preset threshold can be flexibly set based on experience. When the preset threshold is set too large, the pedestrian recognition effect of the trained pedestrian re-identification network may not be good enough, and when the preset threshold is set too small, the function value of the loss function may be difficult to converge during training.

Optionally, the value range of the above preset threshold is [0, 0.01].

Specifically, the value of the preset threshold may be 0.01.

In combination with the first aspect, in certain implementations of the first aspect, the function value of the above-mentioned loss function is less than or equal to a preset threshold, including: the first difference is less than the first preset threshold, and the second difference is less than the second preset threshold.

The above-mentioned first preset threshold and second preset threshold can also be determined based on experience. When the first preset threshold and the second preset threshold are set too large, the pedestrian recognition effect of the trained pedestrian re-identification network may not be good enough, and when the first preset threshold and the second preset threshold are set too small, the function value of the loss function may be difficult to converge during training.

Optionally, the value range of the first preset threshold is [0, 0.4].

Optionally, the value range of the first preset threshold is [0, 0.4].

Specifically, the first preset threshold and the second threshold may both be 0.1.

In combination with the first aspect, in certain implementations of the first aspect, the M training images are training images from multiple image capturing devices, wherein the labeling data of the training images from different image capturing devices are obtained by separate labeling.

That is to say, the images of each image capturing device can be marked separately without considering whether the same pedestrian appears between different image capturing devices. Specifically, if multiple images captured by image capturing device A include pedestrian X, then after marking the M training images captured by image capturing device A, there is no need to search for images containing pedestrian X from images captured by other image capturing devices. This avoids the process of searching for the same pedestrian in images captured by different image capturing devices, saves a lot of marking time, and reduces the complexity of annotation.

In a second aspect, a pedestrian re-identification method is provided, which includes: obtaining an image to be identified; processing the image to be identified using a pedestrian re-identification network to obtain a feature vector of the image to be identified, wherein the pedestrian re-identification network is trained according to the training method of the first aspect above; comparing the feature vector of the image to be identified with the feature vector of an existing pedestrian image to obtain a recognition result of the image to be identified.

In the present application, since the pedestrian re-identification network trained by the training method of the first aspect can better extract features, the pedestrian re-identification network trained by the training method of the first aspect can achieve better pedestrian recognition results for pedestrian recognition.

In combination with the second aspect, in certain implementations of the second aspect, the feature vector of the image to be identified is compared with the feature vector of the existing pedestrian image to obtain the recognition result of the image to be identified, including: outputting the target pedestrian image and the attribute information of the target pedestrian image.

The target pedestrian image may be a pedestrian image whose feature vector is most similar to the feature vector of the image to be identified among the existing pedestrian images, and the attribute information of the target pedestrian image includes the shooting time and shooting location of the target pedestrian image. In addition, the attribute information of the target pedestrian image may also include the identity information of the pedestrian.

In a third aspect, a training device for a person re-identification network is provided, and the training device for a person re-identification network includes modules for executing the method in the first aspect.

In a fourth aspect, a pedestrian re-identification device is provided, which includes various modules for executing the method in the second aspect.

In a fifth aspect, a training device for a pedestrian re-identification network is provided, the device comprising: a memory for storing programs; a processor for executing the programs stored in the memory, when the programs stored in the memory are executed, the processor is used to execute the method in the above-mentioned first aspect.

In a sixth aspect, a pedestrian re-identification device is provided, which includes: a memory for storing programs; a processor for executing the programs stored in the memory, and when the program stored in the memory is executed, the processor is used to execute the method in the above-mentioned second aspect.

In a seventh aspect, a computer device is provided, which includes a training device for a person re-identification network in the third aspect.

In the seventh aspect above, the computer device may specifically be a server or a cloud device, etc.

In an eighth aspect, an electronic device is provided, which includes the pedestrian re-identification device according to the fourth aspect.

In the eighth aspect mentioned above, the electronic device may specifically be a mobile terminal (eg, a smart phone), a tablet computer, a laptop computer, an augmented reality/virtual reality device, a vehicle-mounted terminal device, and the like.

In a ninth aspect, a computer-readable storage medium is provided, wherein the computer-readable storage medium stores a program code, wherein the program code includes instructions for executing the steps in any one of the methods in the first aspect or the second aspect.

In a tenth aspect, a computer program product comprising instructions is provided, and when the computer program product is run on a computer, the computer is caused to execute any one of the methods in the first aspect or the second aspect above.

In the eleventh aspect, a chip is provided, comprising a processor and a data interface, wherein the processor reads instructions stored in a memory through the data interface to execute any one of the methods in the first or second aspect above.

Optionally, as an implementation method, the chip may also include a memory, in which instructions are stored, and the processor is used to execute the instructions stored in the memory. When the instructions are executed, the processor is used to execute any one of the methods in the first aspect or the second aspect above.

The above chip can specifically be a field programmable gate array FPGA or an application specific integrated circuit ASIC.

It should be understood that in the present application, the method of the first aspect may specifically refer to the method in the first aspect and any one of the various implementations of the first aspect, and the method of the second aspect may specifically refer to the method in the second aspect and any one of the various implementations of the second aspect.

BRIEF DESCRIPTION OF THE DRAWINGS

FIG1 is a schematic diagram of the structure of the system architecture provided in an embodiment of the present application;

FIG2 is a schematic diagram of performing pedestrian re-identification using a convolutional neural network model provided in an embodiment of the present application;

FIG3 is a schematic diagram of a chip hardware structure provided in an embodiment of the present application;

FIG4 is a schematic diagram of a system architecture provided in an embodiment of the present application;

FIG5 is a schematic diagram of a possible application scenario of an embodiment of the present application;

FIG6 is a schematic diagram of the overall flow of a training method for a pedestrian re-identification network according to an embodiment of the present application;

FIG7 is a schematic flow chart of a method for training a person re-identification network according to an embodiment of the present application;

FIG8 is a schematic diagram of a process for determining a function value of a loss function;

FIG9 is a schematic flow chart of a pedestrian re-identification method according to an embodiment of the present application;

FIG10 is a schematic block diagram of a training device for a person re-identification network according to an embodiment of the present application;

FIG11 is a schematic block diagram of a training device for a person re-identification network according to an embodiment of the present application;

FIG12 is a schematic block diagram of a pedestrian re-identification device according to an embodiment of the present application;

FIG. 13 is a schematic block diagram of a pedestrian re-identification device according to an embodiment of the present application.

Detailed ways

The following will describe the technical solutions in the embodiments of the present application in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.

The solution of this application can be applied in the fields of urban monitoring, safe city, etc.

Specifically, the present application can be applied in a scenario where an intelligent monitoring system is used to find a person, and the application in this scenario is introduced below.

Intelligent monitoring system to find people:

Taking the intelligent monitoring system deployed in a certain park as an example, the intelligent monitoring system can collect images of pedestrians captured by various image capture devices to form an image library. Next, the images in the image library can be used to train the pedestrian re-identification network (also called the pedestrian re-identification model) to obtain a trained pedestrian re-identification network.

Next, the trained person re-identification network can be used to extract the feature vector of the collected pedestrian image. When a person's whereabouts are suspicious, or there are other situations where the pedestrian needs to be tracked across lenses, the feature vector of the pedestrian image collected by the pedestrian re-identification network can be compared with the feature vector of the image in the image library, and the pedestrian image with the most similar feature vector can be returned, and basic information such as the shooting time and location of these images can be given. After subsequent verification and screening, the person search process can be completed.

In the present application, the pedestrian re-identification network can be a neural network (model). In order to better understand the present application, the relevant terms and concepts of neural networks are introduced below.

(1) Neural Network

The neural network may be composed of neural units, and the neural unit may refer to an operation unit with x _s and intercept 1 as input, and the output of the operation unit may be shown as formula (1):

Wherein, s=1, 2, ...n, n is a natural number greater than 1, _Ws is the weight of _xs , and b is the bias of the neural unit. f is the activation function of the neural unit, which is used to perform nonlinear transformation on the features in the neural network, thereby converting the input signal in the neural unit into the output signal. The output signal of the activation function can be used as the input of the next convolutional layer, and the activation function can be a sigmoid function. A neural network is a network formed by connecting multiple single neural units mentioned above, that is, the output of one neural unit can be the input of another neural unit. The input of each neural unit can be connected to the local receptive field of the previous layer to extract the features of the local receptive field. The local receptive field can be an area composed of several neural units.

(2) Deep Neural Networks

A deep neural network (DNN), also known as a multi-layer neural network, can be understood as a neural network with multiple hidden layers. According to the position of different layers, the neural network inside the DNN can be divided into three categories: input layer, hidden layer, and output layer. Generally speaking, the first layer is the input layer, the last layer is the output layer, and the layers in between are all hidden layers. The layers are fully connected, that is, any neuron in the i-th layer must be connected to any neuron in the i+1-th layer.

Although DNN looks complicated, the work of each layer is not complicated. In simple terms, it can be expressed as the following linear relationship: Among them,/> is the input vector, /> is the output vector, /> is the offset vector, W is the weight matrix (also called coefficient), and α() is the activation function. Each layer is just an input vector/> After such a simple operation, the output vector is obtained/> Since DNN has many layers, coefficient W and offset vector/> The number of these parameters is also relatively large. The definitions of these parameters in DNN are as follows: Taking the coefficient W as an example, assuming that in a three-layer DNN, the linear coefficient from the 4th neuron in the second layer to the 2nd neuron in the third layer is defined as/> The superscript 3 represents the layer number of coefficient W, while the subscripts correspond to the third layer index 2 of the output and the second layer index 4 of the input.

In summary, the coefficients from the kth neuron in the L-1th layer to the jth neuron in the Lth layer are defined as

It should be noted that the input layer does not have a W parameter. In a deep neural network, more hidden layers allow the network to better describe complex situations in the real world. Theoretically, the more parameters a model has, the higher its complexity and the greater its "capacity", which means it can complete more complex learning tasks. Training a deep neural network is the process of learning the weight matrix, and its ultimate goal is to obtain the weight matrix of all layers of the trained deep neural network (a weight matrix formed by many layers of vectors W).

(3) Convolutional Neural Network

Convolutional neural network (CNN) is a deep neural network with a convolutional structure. Convolutional neural network contains a feature extractor consisting of a convolution layer and a subsampling layer, which can be regarded as a filter. Convolutional layer refers to the neuron layer in the convolutional neural network that performs convolution processing on the input signal. In the convolutional layer of the convolutional neural network, a neuron can only be connected to some neurons in the adjacent layers. A convolutional layer usually contains several feature planes, each of which can be composed of some rectangularly arranged neural units. The neural units in the same feature plane share weights, and the shared weights here are convolution kernels. Shared weights can be understood as the way to extract image information is independent of position. The convolution kernel can be initialized in the form of a matrix of random size, and the convolution kernel can obtain reasonable weights through learning during the training process of the convolutional neural network. In addition, the direct benefit of shared weights is to reduce the connection between the layers of the convolutional neural network, while reducing the risk of overfitting.

(4) Residual Network

The residual network is a deep convolutional network proposed in 2015. Compared with traditional convolutional neural networks, the residual network is easier to optimize and can improve accuracy by increasing the depth. The core of the residual network is to solve the side effects (degradation problem) brought by increasing the depth, so that the network performance can be improved by simply increasing the network depth. The residual network generally contains many sub-modules with the same structure. Usually, a residual network (ResNet) is used to connect a number to indicate the number of times the sub-module is repeated. For example, ResNet50 means that there are 50 sub-modules in the residual network.

(6) Classifier

Many neural network structures have a classifier at the end to classify objects in the image. The classifier is generally composed of a fully connected layer and a softmax function (which can be called a normalized exponential function), which can output the probability of different categories based on the input.

(7) Loss function

In the process of training a deep neural network, because we hope that the output of the deep neural network is as close as possible to the value we really want to predict, we can compare the predicted value of the current network with the target value we really want, and then update the weight vector of each layer of the neural network according to the difference between the two (of course, there is usually an initialization process before the first update, that is, pre-configuring parameters for each layer in the deep neural network). For example, if the predicted value of the network is high, adjust the weight vector to make it predict a lower value, and keep adjusting until the deep neural network can predict the target value we really want or a value very close to the target value we really want. Therefore, it is necessary to pre-define "how to compare the difference between the predicted value and the target value", which is the loss function or objective function, which are important equations used to measure the difference between the predicted value and the target value. Among them, taking the loss function as an example, the higher the output value (loss) of the loss function, the greater the difference, so the training of the deep neural network becomes a process of minimizing this loss as much as possible.

(8) Back propagation algorithm

Neural networks can use the error back propagation (BP) algorithm to correct the values of the parameters in the initial neural network model during the training process, so that the reconstruction error loss of the neural network model becomes smaller and smaller. Specifically, the forward transmission of the input signal to the output will generate error loss, and the parameters in the initial neural network model are updated by back propagating the error loss information, so that the error loss converges. The back propagation algorithm is a back propagation movement dominated by error loss, which aims to obtain the optimal parameters of the neural network model, such as the weight matrix.

The system architecture of the embodiment of the present application is described in detail below in conjunction with FIG. 1 .

FIG1 is a schematic diagram of a system architecture of an embodiment of the present application. As shown in FIG1 , the system architecture 100 includes an execution device 110 , a training device 120 , a database 130 , a client device 140 , a data storage system 150 , and a data acquisition system 160 .

In addition, the execution device 110 includes a calculation module 111, an I/O interface 112, a preprocessing module 113 and a preprocessing module 114. The calculation module 111 may include the target model/rule 101, and the preprocessing module 113 and the preprocessing module 114 are optional.

The data acquisition device 160 is used to collect training data. For the training method of the pedestrian re-identification network of the embodiment of the present application, the training data may include M training images and the annotation data of the M training images. After collecting the training data, the data acquisition device 160 stores the training data in the database 130, and the training device 120 trains the target model/rule 101 based on the training data maintained in the database 130.

The following describes how the training device 120 obtains the target model/rule 101 based on the training data. The training device 120 performs feature extraction on the input training image to obtain a feature vector of the training image, and repeatedly performs feature extraction on the input training image until the function value of the loss function meets the preset requirements (less than or equal to the preset threshold), thereby completing the training of the target model/rule 101.

It should be understood that the training of the above target model/rule 101 can be an unsupervised training.

The above-mentioned target model/rule 101 can be used to implement the pedestrian re-identification method of the embodiment of the present application, that is, by inputting a pedestrian image (the pedestrian image may be an image that needs to be recognized as a pedestrian) into the target model/rule 101, a feature vector extracted from the pedestrian image can be obtained, and pedestrian recognition is performed based on the extracted feature vector to determine the recognition result of the pedestrian. The target model/rule 101 in the embodiment of the present application can specifically be a neural network. It should be noted that in actual applications, the training data maintained in the database 130 does not necessarily all come from the data acquisition device 160, but may also be received from other devices. It should also be noted that the training device 120 does not necessarily train the target model/rule 101 entirely based on the training data maintained by the database 130, and it is also possible to obtain training data from the cloud or other places for model training. The above description should not be used as a limitation on the embodiments of the present application.

The target model/rule 101 obtained by training the training device 120 can be applied to different systems or devices, such as the execution device 110 shown in FIG1 . The execution device 110 can be a terminal, such as a mobile phone terminal, a tablet computer, a laptop computer, an augmented reality (AR)/virtual reality (VR), a vehicle terminal, etc., and can also be a server or a cloud. In FIG1 , the execution device 110 is configured with an input/output (I/O) interface 112 for data interaction with an external device. The user can input data to the I/O interface 112 through the client device 140. The input data in the embodiment of the present application may include: a pedestrian image input by the client device. The client device 140 here can specifically be a monitoring device.

The preprocessing module 113 and the preprocessing module 114 are used to preprocess the input data (such as pedestrian images) received by the I/O interface 112. In the embodiment of the present application, there may be no preprocessing module 113 and the preprocessing module 114 or only one preprocessing module. When there is no preprocessing module 113 and the preprocessing module 114, the computing module 111 may be directly used to process the input data.

When the execution device 110 preprocesses the input data, or when the computing module 111 of the execution device 110 performs calculations and other related processing, the execution device 110 can call the data, code, etc. in the data storage system 150 for corresponding processing, and can also store the data, instructions, etc. obtained from the corresponding processing into the data storage system 150.

Finally, the I/O interface 112 presents the processing result (specifically, a high-quality image obtained by pedestrian re-identification), such as the recognition result of the image to be recognized obtained by performing pedestrian re-identification processing on the pedestrian image by the target model/rule 101, to the client device 140, thereby providing it to the user.

Specifically, the high-quality image obtained by re-identifying the pedestrian through the target model/rule 101 in the computing module 111 can be processed by the pre-processing module 113 (and can also be processed by the pre-processing module 114) (for example, image rendering processing) and the processing result can be sent to the I/O interface, and then the I/O interface can send the processing result to the client device 140 for display.

It should be understood that when the pre-processing module 113 and the pre-processing module 114 do not exist in the above-mentioned system architecture 100, the computing module 111 can also transmit the high-quality image obtained through the pedestrian re-identification processing to the I/O interface, and then the I/O interface sends the processing result to the client device 140 for display.

It is worth noting that the training device 120 can target different goals or different tasks (for example, the training device can be trained on real high-quality images and approximate low-quality images in different scenarios), and generate corresponding target models/rules 101 based on different training data. The corresponding target models/rules 101 can be used to achieve the above goals or complete the above tasks, thereby providing users with the desired results.

It is worth noting that Figure 1 is only a schematic diagram of a system architecture provided by an embodiment of the present application. The positional relationship between the devices, components, modules, etc. shown in the figure does not constitute any limitation. For example, in Figure 1, the data storage system 150 is an external memory relative to the execution device 110. In other cases, the data storage system 150 can also be placed in the execution device 110.

As shown in Fig. 1, the target model/rule 101 obtained by training with the training device 120 may be a neural network (model). Specifically, the neural network (model) may be a CNN or a deep convolutional neural network (DCNN).

Since CNN is a very common neural network, the following focuses on the detailed introduction of the structure of CNN in conjunction with Figure 2. As mentioned in the basic concept introduction above, convolutional neural network is a deep neural network with a convolution structure and a deep learning architecture. A deep learning architecture refers to multiple levels of learning at different abstract levels through machine learning algorithms. As a deep learning architecture, CNN is a feed-forward artificial neural network in which each neuron can respond to the image input into it.

As shown in Fig. 2, a convolutional neural network (CNN) 200 may include an input layer 210, a convolutional layer/pooling layer 220 (wherein the pooling layer is optional), and a fully connected layer 230. The relevant contents of these layers are described in detail below.

Convolutional layer/pooling layer 220:

Convolutional Layer:

As shown in FIG2 , the convolution layer/pooling layer 220 may include layers 221-226, for example: in one implementation, layer 221 is a convolution layer, layer 222 is a pooling layer, layer 223 is a convolution layer, layer 224 is a pooling layer, layer 225 is a convolution layer, and layer 226 is a pooling layer; in another implementation, layers 221 and 222 are convolution layers, layer 223 is a pooling layer, layers 224 and 225 are convolution layers, and layer 226 is a pooling layer. That is, the output of a convolution layer can be used as the input of a subsequent pooling layer, or as the input of another convolution layer to continue the convolution operation.

The following will take the convolution layer 221 as an example to introduce the internal working principle of a convolution layer.

The convolution layer 221 may include a plurality of convolution operators, which are also called kernels. The convolution operator is equivalent to a filter that extracts specific information from the input image matrix in image processing. The convolution operator can be essentially a weight matrix, which is usually predefined. In the process of performing convolution operations on the image, the weight matrix is usually processed one pixel after another (or two pixels after two pixels... depending on the value of the stride) in the horizontal direction on the input image, thereby completing the work of extracting specific features from the image. The size of the weight matrix should be related to the size of the image. It should be noted that the depth dimension of the weight matrix is the same as the depth dimension of the input image. In the process of performing convolution operations, the weight matrix will extend to the entire depth of the input image. Therefore, convolution with a single weight matrix will produce a convolution output with a single depth dimension, but in most cases, a single weight matrix is not used, but multiple weight matrices of the same size (row × column), that is, multiple isotype matrices, are applied. The output of each weight matrix is stacked to form the depth dimension of the convolution image, and the dimension here can be understood as being determined by the "multiple" mentioned above. Different weight matrices can be used to extract different features in the image, for example, one weight matrix is used to extract image edge information, another weight matrix is used to extract specific colors of the image, and another weight matrix is used to blur unnecessary noise points in the image, etc. The multiple weight matrices have the same size (rows × columns), and the convolution feature maps extracted by the multiple weight matrices of the same size are also the same size. The extracted multiple convolution feature maps of the same size are then merged to form the output of the convolution operation.

The weight values in these weight matrices need to be obtained through a lot of training in practical applications. The weight matrices formed by the weight values obtained through training can be used to extract information from the input image, so that the convolutional neural network 200 can make correct predictions.

When the convolutional neural network 200 has multiple convolutional layers, the initial convolutional layer (for example, 221) often extracts more general features, which can also be called low-level features. As the depth of the convolutional neural network 200 increases, the features extracted by the later convolutional layers (for example, 226) become more and more complex, such as high-level semantic features. Features with higher semantics are more suitable for the problem to be solved.

Pooling layer:

Since it is often necessary to reduce the number of training parameters, it is often necessary to periodically introduce a pooling layer after the convolution layer. In each layer 221-226 as shown in 220 in FIG. 2, a convolution layer may be followed by a pooling layer, or multiple convolution layers may be followed by one or more pooling layers. In the image processing process, the only purpose of the pooling layer is to reduce the spatial size of the image. The pooling layer may include an average pooling operator and/or a maximum pooling operator to sample the input image to obtain an image of smaller size. The average pooling operator may calculate the pixel values in the image within a specific range to generate an average value as the result of average pooling. The maximum pooling operator may take the pixel with the largest value in the range within a specific range as the result of maximum pooling. In addition, just as the size of the weight matrix used in the convolution layer should be related to the image size, the operator in the pooling layer should also be related to the image size. The size of the image output after processing by the pooling layer may be smaller than the size of the image input to the pooling layer, and each pixel in the image output by the pooling layer represents the average value or maximum value of the corresponding sub-region of the image input to the pooling layer.

Fully connected layer 230:

After being processed by the convolution layer/pooling layer 220, the convolution neural network 200 is not sufficient to output the required output information. Because as mentioned above, the convolution layer/pooling layer 220 will only extract features and reduce the parameters brought by the input image. However, in order to generate the final output information (the required class information or other related information), the convolution neural network 200 needs to use the fully connected layer 230 to generate one or a group of outputs of the required number of classes. Therefore, the fully connected layer 230 may include multiple hidden layers (231, 232 to 23n as shown in Figure 2) and an output layer 240. The parameters contained in the multiple hidden layers can be pre-trained according to the relevant training data of the specific task type. For example, the task type may include image recognition, image classification, image super-resolution reconstruction, etc.

After the multiple hidden layers in the fully connected layer 230, that is, the last layer of the entire convolutional neural network 200 is the output layer 240, which has a loss function similar to the classification cross entropy, which is specifically used to calculate the prediction error. Once the forward propagation of the entire convolutional neural network 200 (as shown in FIG. 2, the propagation from 210 to 240 is the forward propagation) is completed, the back propagation (as shown in FIG. 2, the propagation from 240 to 210 is the back propagation) will begin to update the weight values and biases of the aforementioned layers to reduce the loss of the convolutional neural network 200 and the error between the result output by the convolutional neural network 200 through the output layer and the ideal result.

It should be noted that the convolutional neural network 200 shown in FIG. 2 is only an example of a convolutional neural network. In specific applications, the convolutional neural network may also exist in the form of other network models.

It should be understood that the convolutional neural network (CNN) 200 shown in Figure 2 can be used to perform the pedestrian re-identification method of the embodiment of the present application. As shown in Figure 2, after the pedestrian image is processed by the input layer 210, the convolution layer/pooling layer 220 and the fully connected layer 230, the image features of the image to be identified can be obtained, and subsequently the recognition result of the image to be identified can be obtained based on the image features of the image to be identified.

FIG3 is a chip hardware structure provided in an embodiment of the present application, and the chip includes a neural network processor 50. The chip can be set in the execution device 110 as shown in FIG1 to complete the calculation work of the calculation module 111. The chip can also be set in the training device 120 as shown in FIG1 to complete the training work of the training device 120 and output the target model/rule 101. The algorithms of each layer in the convolutional neural network shown in FIG2 can be implemented in the chip shown in FIG3.

The neural-network processing unit (NPU) 50 is mounted on the host central processing unit (CPU) as a coprocessor, and the host CPU assigns tasks. The core part of the NPU is the operation circuit 503, and the controller 504 controls the operation circuit 503 to extract data from the memory (weight memory or input memory) and perform operations.

In some implementations, the operation circuit 503 includes multiple processing units (process engines, PEs) inside. In some implementations, the operation circuit 503 is a two-dimensional systolic array. The operation circuit 503 can also be a one-dimensional systolic array or other electronic circuits capable of performing mathematical operations such as multiplication and addition. In some implementations, the operation circuit 503 is a general-purpose matrix processor.

For example, assume there is an input matrix A, a weight matrix B, and an output matrix C. The operation circuit 503 takes the corresponding data of the matrix B from the weight memory 502 and caches it on each PE in the operation circuit 503. The operation circuit 503 takes the matrix A data from the input memory 501 and performs a matrix operation with the matrix B, and the partial result or the final result of the matrix is stored in the accumulator 508.

The vector calculation unit 507 can further process the output of the operation circuit 503, such as vector multiplication, vector addition, exponential operation, logarithmic operation, size comparison, etc. For example, the vector calculation unit 507 can be used for network calculations of non-convolutional/non-FC layers in a neural network, such as pooling, batch normalization, local response normalization, etc.

In some implementations, the vector calculation unit can 507 store the processed output vector to the unified buffer 506. For example, the vector calculation unit 507 can apply a nonlinear function to the output of the operation circuit 503, such as a vector of accumulated values, to generate an activation value. In some implementations, the vector calculation unit 507 generates a normalized value, a merged value, or both. In some implementations, the processed output vector can be used as an activation input to the operation circuit 503, such as for use in a subsequent layer in a neural network.

The unified memory 506 is used to store input data and output data.

The weight data is directly transferred from the external memory to the input memory 501 and/or the unified memory 506 through the direct memory access controller 505 (DMAC), the weight data in the external memory is stored in the weight memory 502, and the data in the unified memory 506 is stored in the external memory.

The bus interface unit (BIU) 510 is used to implement the interaction between the main CPU, DMAC and instruction fetch memory 509 through the bus.

An instruction fetch buffer 509 connected to the controller 504 and used to store instructions used by the controller 504;

The controller 504 is used to call the instructions cached in the memory 509 to control the working process of the computing accelerator.

Generally, the unified memory 506, the input memory 501, the weight memory 502 and the instruction fetch memory 509 are all on-chip memories, and the external memory is a memory outside the NPU, which can be a double data rate synchronous dynamic random access memory (DDR SDRAM), a high bandwidth memory (HBM) or other readable and writable memory.

In addition, in the present application, the operations of each layer in the convolutional neural network shown in Figure 2 can be performed by the operation circuit 503 or the vector calculation unit 507.

As shown in Fig. 4, the embodiment of the present application provides a system architecture 300. The system architecture includes a local device 301, a local device 302, an execution device 210 and a data storage system 250, wherein the local device 301 and the local device 302 are connected to the execution device 210 via a communication network.

The execution device 210 can be implemented by one or more servers. Optionally, the execution device 210 can be used in conjunction with other computing devices, such as data storage devices, routers, load balancers, and other devices. The execution device 210 can be arranged at one physical site, or distributed at multiple physical sites. The execution device 210 can use the data in the data storage system 250, or call the program code in the data storage system 250 to implement the pedestrian re-identification method of the embodiment of the present application.

Users can operate their respective user devices (e.g., local device 301 and local device 302) to interact with execution device 210. Each local device can represent any computing device, such as a personal computer, a computer workstation, a smart phone, a tablet computer, a smart camera, a smart car or other type of cellular phone, a media consumption device, a wearable device, a set-top box, a game console, etc.

The local device of each user can interact with the execution device 210 through a communication network of any communication mechanism/communication standard. The communication network can be a wide area network, a local area network, a point-to-point connection, etc., or any combination thereof.

In one implementation, the local device 301 and the local device 302 obtain relevant parameters of the target neural network from the execution device 210, deploy the target neural network on the local device 301 and the local device 302, and use the target neural network to perform pedestrian re-identification.

In another implementation, the target neural network can be directly deployed on the execution device 210, and the execution device 210 obtains pedestrian images from the local device 301 and the local device 302 (the local device 301 and the local device 302 can upload the pedestrian images to the execution device 210), and performs pedestrian re-identification on the pedestrian images according to the target neural network, and sends the high-quality images obtained by pedestrian re-identification to the local device 301 and the local device 302.

The execution device 210 may also be referred to as a cloud device. In this case, the execution device 210 is generally deployed in the cloud.

FIG. 5 is a schematic diagram of a possible application scenario of an embodiment of the present application.

As shown in FIG5 , in the present application, the pedestrian re-identification network can be trained by annotating data of a single image shooting device to obtain a trained pedestrian re-identification network, which can process pedestrian images to obtain feature vectors of pedestrian images. Next, by performing feature comparison between the feature vectors of the pedestrian image and feature vectors in the image library, the person to be found can be obtained. Specifically, the target pedestrian image that is most similar to the feature vector of the pedestrian image can be found through feature comparison, and basic information such as the shooting time and location of the target pedestrian image can be output.

It should be understood that the image library stores the feature vectors of each pedestrian image and the relevant information of the pedestrian corresponding to the pedestrian image.

It should be understood that the single image capture device annotation data here may include multiple training images and annotation data of multiple training images. Single image capture device annotation data refers to individual annotation for each training image acquired by the image capture device, without the need to search for the same pedestrian between different image capture devices. This annotation method does not need to pay attention to the relationship between the training images captured by different image capture devices, which can save a lot of labeling time and reduce the complexity of annotation. The above-mentioned multiple training images and the annotation data of multiple training images can also be collectively referred to as training data.

FIG6 is a schematic diagram of the overall flow of a training method for a person re-identification network according to an embodiment of the present application.

As shown in Figure 6, by individually annotating the video images acquired by each image capture device, single image capture device annotated data can be obtained. The biggest advantage of single image capture device annotated data is that it is easy to annotate and collect. In this application, single image capture device annotated data does not require the same pedestrian to appear under multiple image capture devices.

In the single image capture device annotation data, it is assumed that each pedestrian only appears in one image capture device (or one image capture device group). In this way, after obtaining the pedestrian image in the video using pedestrian detection and tracking, only a small amount of manpower is needed to associate several images of the same person in similar frames to form annotations. Moreover, the annotations of each image capture device are relatively independent, and there will be no overlap in the pedestrian numbers on different image capture devices. By setting different acquisition time periods for different image capture devices, the number of people who appear repeatedly in the video captured by each image capture device can be reduced, thereby meeting the requirements of single image capture device annotation data.

In some relatively small scenes (for example, an office park), most people have a small range of activities, and quite a few people only appear in a certain image capture device group. Such data can naturally meet this requirement. Since the cameras in an image capture device group have similar or overlapping fields of view and similar lighting conditions, these cameras can basically be equivalent to one camera.

After obtaining the single image shooting device annotated data, the single image shooting device annotated data can be used to train the pedestrian re-identification network (model), and the trained pedestrian re-identification network can be used for testing and deployment. Specifically, the trained pedestrian re-identification network can be used to execute the pedestrian re-identification method of the embodiment of the present application.

FIG7 is a schematic flow chart of a training method for a person re-identification network according to an embodiment of the present application. The method shown in FIG7 can be performed by a training device for a person re-identification network according to an embodiment of the present application (for example, it can be performed by the devices shown in FIG10 and FIG11 ). The method shown in FIG7 includes steps 1001 to 1008, which are described in detail below.

1001. Start.

Step 1001 indicates starting the training process of the person re-identification network.

1002. Obtain training data.

The training data in the above step 1002 includes M (M is an integer greater than 1) training images and annotation data of the M training images, wherein, in the M training images, each training image includes a pedestrian, and the annotation data of each training image includes a bounding box and pedestrian identification information of the pedestrian in each training image, different pedestrians correspond to different pedestrian identification information, and in the M training images, the training images with the same pedestrian identification information come from the same image capture device.

The above-mentioned image capturing device may specifically be a device such as a video camera or a still camera that can capture images of pedestrians.

The pedestrian identification information in the above step 1002 can also be called pedestrian identity identification information, which is a kind of information used to indicate the identity of the pedestrian. Each pedestrian can correspond to unique pedestrian identification information. There are many ways to represent the pedestrian identification information, as long as it can indicate the identity information of the pedestrian. For example, the pedestrian identification information can specifically be the pedestrian identity (identity, ID), that is, a unique ID can be assigned to each pedestrian.

1003. Initialize the network parameters of the pedestrian re-identification network to obtain initial values of the network parameters of the pedestrian re-identification network.

In the above step 1003, the network parameters of the pedestrian re-identification network can be randomly set to obtain the initial values of the network parameters of the pedestrian re-identification network.

1004. Input a batch of training images from the M training images into a person re-identification network for feature extraction to obtain a feature vector for each training image in the batch of training images.

The above-mentioned batch of training images is part of the M training images. When the M training images are used to train the pedestrian re-recognition network, the M training images can be divided into different batches to train the pedestrian re-recognition network. The number of training images in each batch can be the same or different.

For example, if there are 5,000 training images in total, 100 training images can be input in each batch to train the pedestrian re-identification network.

The above-mentioned batch of training images may include N anchor images, wherein the N anchor images are any N training images in the above-mentioned batch of training images, each anchor image in the N anchor images corresponds to a most difficult positive sample image, a first most difficult negative sample image and a second most difficult negative sample image, N is a positive integer, and N is less than M.

The most difficult positive sample image, the first most difficult negative sample image and the second most difficult negative sample image corresponding to each anchor point image are described below.

The most difficult positive sample image corresponding to each anchor image: the training image in the above batch of training images that has the same pedestrian identification information as each anchor image and has the farthest distance from the feature vector of each anchor image;

The first most difficult negative sample image corresponding to each anchor image: a training image in the above batch of training images that comes from the same image capture device as each anchor image, has different pedestrian identification information from each anchor image, and has the closest distance to the feature vector of each anchor image;

The second most difficult negative sample image corresponding to each anchor image: a training image in the above batch of training images that comes from a different image capture device than each anchor image, has different pedestrian identification information from each anchor image, and has the closest distance to the feature vector of each anchor image.

1005. Determine a function value of a loss function according to the feature vectors of the above batch of training images.

The function value of the loss function in the above step 1005 is obtained by averaging the function values of the N first loss functions.

The function value of each of the N first loss functions is calculated based on the first difference and the second difference corresponding to each of the N anchor point images.

Optionally, the function value of each of the first loss functions is the sum of the first difference and the second difference corresponding to each anchor point image.

The above N is a positive integer, and the above N is less than M. When N=1, there is only one function value of the first loss function, and at this time, the function value of the first loss function can be directly used as the function value of the loss function in step 1005 .

For example, the function value of the first loss function may be as shown in formula (2).

L1＝D1+D2 (2)

Among them, L1 represents the function value of the first loss function, D1 represents the above-mentioned first difference, and D2 represents the above-mentioned second difference.

Optionally, the function value of each of the above-mentioned first loss functions is the sum of the absolute value of the first difference and the absolute value of the second difference corresponding to each anchor point image.

For example, the function value of the first loss function may be as shown in formula (3).

L1＝|D1|+|D2| (3)

Wherein, L1 represents the function value of the first loss function, |D1| represents the absolute value of the first difference, and |D2| represents the absolute value of the second difference.

Optionally, the function value of each of the above-mentioned first loss functions is the sum of the first difference, the second difference and other constant terms corresponding to each anchor point image.

For example, the function value of the first loss function may be as shown in formula (4).

L1＝D1+D2+m (4)

Among them, L1 represents the function value of the first loss function, D1 represents the above-mentioned first difference, D2 represents the above-mentioned second difference, m represents a constant, and the size of m can be set to a suitable value based on experience.

For another example, the function value of the first loss function can be shown as formula (5).

L1＝| _m1 +D1|+| _m2 +D2| (5)

Among them, L1 represents the function value of the first loss function, |D1| represents the above-mentioned first difference, |D2| represents the above-mentioned second difference, _m1 and _m2 represent constants, and the sizes of _m1 and _m2 can be set to appropriate values based on experience.

It should be understood that in the above text, obtaining the absolute values of D1 and D2 when calculating the function value of the first loss function is only an optional implementation method. In fact, other operations can be performed on D1 and D2 when determining the function value of the first loss function. For example, the [X] ₊ operation can be performed on D1 and D2 (this operation can be called an operation of taking the positive part of the function value).

When X is greater than 0, [X] ₊ = X, and when X is less than 0, [X] ₊ = 0. (For details, please refer to https://en.wikipedia.org/wiki/Positive_and_negative_parts)

The meanings of the first difference and the second difference are explained below.

The first difference value corresponding to each anchor image: the difference between the most difficult positive sample distance corresponding to each anchor image and the second most difficult negative sample distance corresponding to each anchor image;

The second difference value corresponding to each anchor image: the difference between the second most difficult negative sample distance corresponding to each anchor image and the first most difficult negative sample distance corresponding to each anchor image.

The first difference and the second difference are obtained by subtracting different distances. The meanings of these distances are explained below.

The most difficult positive sample distance corresponding to each anchor image: the distance between the feature vector of the most difficult positive sample image corresponding to each anchor image and the feature vector of each anchor image;

The second most difficult negative sample distance corresponding to each anchor image: the distance between the feature vector of the second most difficult negative sample image corresponding to each anchor image and the feature vector of each anchor image;

The first hardest negative sample distance corresponding to each anchor image: the distance between the feature vector of the first hardest negative sample image corresponding to each anchor image and the feature vector of each anchor image.

Specifically, assuming that the number of image capture devices corresponding to a batch of images during training is C, the number of pedestrians under each image capture device is P, and the number of images of each pedestrian is K, then the number of images in a batch is C×P×K. The anchor images in this batch are denoted as Note/> is the feature (vector) output by the network model, let ||f ₁ -f ₂ || be the Euclidean distance between the two features f ₁ and f ₂ , then the above-mentioned most difficult positive sample distance can be shown as formula (6).

The second most difficult negative sample distance can be expressed as formula (7).

The first most difficult negative sample distance can be expressed as formula (8).

The first difference may be the difference between formula (6) and formula (7), and the second difference may be the difference between formula (7) and formula (8).

The loss function composed of the first difference and the second difference can be shown as formula (9). Wherein, L represents the loss function, _m1 and _m2 are two constants, and the specific values can be set according to experience. For example, _m1 = 0.1, _m2 = 0.1.

In the above formula (9), Expressing support for/> Perform [X] ₊ operation, when /> When the value of is greater than or equal to 0, /> The value of is When/> When the value of is less than 0, /> The value of is 0.

In the above formula (9), Expressing support for/> Perform [X] ₊ operation, when /> When the value of is greater than or equal to 0, /> The value of is /> And when/> When it is less than 0, The value of is 0.

1006. Update the network parameters of the person re-identification network according to the function value of the loss function.

Specifically, the network parameters of the person re-identification network can be updated according to the function value of the loss function shown in the above formula (9), and the function value of the loss function shown in the formula (9) is made smaller and smaller during the updating process.

1007. Determine whether the pedestrian re-identification network meets the preset requirements.

Optionally, the pedestrian re-identification network meets preset requirements, including: the pedestrian re-identification network meets at least one of the following conditions:

(1) The pedestrian recognition performance of the pedestrian re-identification network meets the preset performance requirements;

(2) The number of updates of the network parameters of the person re-identification network is greater than or equal to the preset number;

(3) The function value of the loss function is less than or equal to the preset threshold.

In step 1007, when the pedestrian re-identification network satisfies at least one of the above conditions (1) to (3), it can be determined that the pedestrian re-identification network meets the preset requirements, and step 1008 is executed, and the training process of the pedestrian re-identification network ends; when the pedestrian re-identification network does not meet any of the above conditions (1) to (3), it means that the pedestrian re-identification network has not yet met the preset requirements, and it is necessary to continue to train the pedestrian re-identification network, that is, re-execute steps 1004 to 1007 until a pedestrian re-identification network that meets the preset requirements is obtained.

The above-mentioned preset threshold can be flexibly set based on experience. When the preset threshold is set too large, the pedestrian recognition effect of the trained pedestrian re-identification network may not be good enough, and when the preset threshold is set too small, the function value of the loss function may be difficult to converge during training.

Optionally, the value range of the above preset threshold is [0, 0.01].

Specifically, the value of the preset threshold may be 0.01.

The function value of the above loss function is less than or equal to a preset threshold, specifically including: the first difference is less than the first preset threshold, and the second difference is less than the second preset threshold.

The above-mentioned first preset threshold and second preset threshold can also be determined based on experience. When the first preset threshold and the second preset threshold are set too large, the pedestrian recognition effect of the trained pedestrian re-identification network may not be good enough, and when the first preset threshold and the second preset threshold are set too small, the function value of the loss function may be difficult to converge during training.

Optionally, the value range of the first preset threshold is [0, 0.4].

Optionally, the value range of the first preset threshold is [0, 0.4].

Specifically, the first preset threshold and the second threshold may both be 0.1.

1008. Training is over.

In addition, in the present application, several training images coming from the same image capturing device means that these several training images are captured by the same image capturing device.

In the present application, the most difficult negative sample images from different image capturing devices and the same image capturing device are taken into account in the process of constructing the loss function, and the first difference and the second difference are reduced as much as possible during the training process, so as to eliminate the interference of the image capturing device's own information on the image information as much as possible, so that the trained pedestrian re-identification network can more accurately extract features from the image.

Specifically, during the training process of the pedestrian re-identification network, the network parameters of the pedestrian re-identification network are optimized to make the first difference and the second difference as small as possible, thereby making the difference between the most difficult positive sample distance and the second most difficult negative sample distance and the difference between the second most difficult negative sample distance and the first most difficult negative sample distance as small as possible, so that the pedestrian re-identification network can distinguish the features of the most difficult positive image and the second most difficult negative sample image, as well as the features of the second most difficult negative sample image and the first most difficult negative sample image as much as possible, so that the trained pedestrian re-identification network can better and more accurately extract features from the image.

The process of determining the function value of the loss function according to a batch of training images in the above steps 1004 and 1005 is described in detail below in conjunction with FIG. 8 .

As shown in Figure 8, after a batch of training images are input into the person re-identification network, feature vectors of the batch of training images can be obtained. Next, multiple anchor images can be selected from the batch of training images, and the corresponding most difficult positive sample image, the first most difficult negative sample image, and the second most difficult negative sample image can be determined for each of the multiple anchor images.

In this way, many training image groups consisting of four training images (the anchor image, the most difficult positive sample image corresponding to the anchor image, the first most difficult negative sample image corresponding to the anchor image, and the second most difficult negative sample image corresponding to the anchor image) can be obtained, and then a first loss function can be determined according to the distance relationship between the feature vectors of the training images in each training image group.

As shown in FIG8 , there are a total of N training image groups, and a total of N first loss functions can be determined based on the N training image groups. Next, the function values of the N first loss functions are averaged to obtain the function value of the loss function in the above step 1005.

It should be understood that the above N training image groups contain a total of N anchor images, and the N anchor images are different, that is, each training image group corresponds to a unique anchor image. However, other training images (other images except anchor images) contained in different training image groups can be the same. For example, the most difficult positive sample image in the first training image group is the same as the most difficult positive sample image in the second training group.

For another example, assuming that the number of the above batch of training images is 100, then 10 (or other numbers, here only 10 is used as an example) anchor images can be selected from the 100 training images, and then the corresponding most difficult positive sample image, the first most difficult negative sample image and the second most difficult negative sample image are selected for each anchor image from the 100 training images. Thus, 10 training image groups are obtained, and 10 function values of the first loss function can be obtained according to the 10 training image groups. Next, by averaging the function values of the 10 first loss functions, the function value of the loss function in the above step 1005 can be obtained.

The design and training process of the pedestrian re-identification network are introduced in detail below.

The pedestrian re-identification network in this application can adopt the existing residual network (for example, ResNet50) as the network body, remove the last fully connected layer, add a global average pooling layer after the last residual block (ResBlock), and obtain a 2048-dimensional (or other numerical) feature vector as the output of the network model.

In each batch of training images, each camera can capture images of 4 people, and 8 images are collected for each person. If there are fewer than 8 images of one person, the images are repeated to make up 8.

When training the trained person re-identification network, the above formula (9) can be used as the loss function. When testing, the number of cameras of different data sets can be different. For example, for the DukeMTMC-reID data set, there are 8 cameras, and C in formula (9) is 8; for the Market-1501 data set, there are 6 cameras, and C in formula (9) is 6.

The two parameters in the loss function shown in the above formula (9) can be m ₁ = 0.1 and m ₂ = 0.1 respectively. The input training image can be scaled to 256x128 pixels. During training, the adaptive moment estimation (Adam) optimizer can be used to train the network parameters, and the learning rate can be set to 2×10 ^-4 . After 100 rounds of training, the learning rate decays exponentially until the learning rate can be set to 2×10 ^-7 after 200 rounds of learning, at which time the training can be stopped.

The pedestrian re-identification network trained according to the training method of the pedestrian re-identification network of the embodiment of the present application can be used to execute the pedestrian re-identification method of the embodiment of the present application. The pedestrian re-identification method of the embodiment of the present application is described below in conjunction with the accompanying drawings.

FIG9 is a schematic flow chart of a pedestrian re-identification method according to an embodiment of the present application. The pedestrian re-identification method shown in FIG9 can be performed by a pedestrian re-identification device according to an embodiment of the present application (for example, it can be performed by the devices shown in FIG12 and FIG13 ). The pedestrian re-identification method shown in FIG9 includes steps 2001 to 2003, and steps 2001 to 2003 are described in detail below.

2001. Obtain an image to be identified.

2002. The image to be identified is processed using a person re-identification network to obtain a feature vector of the image to be identified.

Among them, the pedestrian re-identification network used in step 2002 can be trained according to the training method of the pedestrian re-identification network of the embodiment of the present application. Specifically, the pedestrian re-identification network in step 2002 can be trained by the method shown in Figure 7.

2003. Compare the feature vector of the image to be identified with the feature vector of the existing pedestrian image to obtain the recognition result of the image to be identified.

In the present application, the pedestrian re-identification network trained by the training method of the pedestrian re-identification network of the embodiment of the present application can better extract features. Therefore, using the pedestrian re-identification network to process the image to be identified can obtain better pedestrian recognition results.

Optionally, the above step 2003 specifically includes: determining to output a target pedestrian image according to comparing a feature vector of the image to be identified with a feature vector of an existing pedestrian image; and outputting the target pedestrian image and attribute information of the target pedestrian image.

The target pedestrian image may be a pedestrian image whose feature vector is most similar to the feature vector of the image to be identified among the existing pedestrian images, and the attribute information of the target pedestrian image includes the shooting time and shooting location of the target pedestrian image. In addition, the attribute information of the target pedestrian image may also include the identity information of the pedestrian.

The effect of pedestrian recognition of the pedestrian re-identification network in the embodiment of the present application is described below in combination with specific test results.

Table 1

Table 1 shows the test results of different schemes on different data sets, where the test results include Rank-1 and mean average precision (mAP). Rank-1 indicates the probability that the image whose feature vector in the existing image is closest to the feature vector of the image to be identified belongs to the same pedestrian as the image to be identified.

In the above Table 1, dataset 1 is Duke-SCT and dataset 2 is Market-SCT.

Duke-SCT is a subset of the DukeMTMC-reID dataset, and Market-SCT is a subset of the Market-1501 dataset. When obtaining Duke-SCT and Market-SCT, we processed the training data from the original datasets (DukeMTMC-reID and Market-1501) as follows: each pedestrian randomly selected an image from a camera and retained it (different pedestrians may select different cameras), thus forming new datasets Duke-SCT and Market-SCT. At the same time, the test set remained unchanged.

Existing solution 1: a discriminative feature learning approach for deep face, which was published at the European Conference on Computer Vision (ECCV) in 2016;

Existing solution 2: deep hypersphere embedding for face recognition, which was published at the 2017 Conference on Computer Vision and Pattern Recognition (CVPR).

Existing solution 3: Additive angular margin loss for deep face recognition, which was published in CVPR in 2019;

Existing solution 4: person retrieval with refined partpooling, which was published in ECCV in 2018;

Existing Solution 5: Part-aligned bilinear representations for person re-identification, which was published in ECCV in 2018;

Existing Solution 6: Learning discriminative features with multiple granularities for person re-Identification, which was published at the Association for Computing Machinery International Conference on Multimedia (ACMMM) in 2018.

As can be seen from Table 1, the Rank-1 and mAP of the proposed solution are superior to the existing solutions in both dataset 1 and dataset 2, and have better recognition effect.

Fig. 10 is a schematic block diagram of a training device for a person re-identification network according to an embodiment of the present application. The training device 8000 for a person re-identification network shown in Fig. 10 comprises an acquisition unit 8001 and a training unit 8002.

The acquisition unit 8001 and the training unit 8002 may be used to execute the training method of the person re-identification network of the embodiment of the present application.

Specifically, the acquisition unit 8001 may execute the above steps 1001 and 1002, and the training unit 8002 may execute the above steps 1003 to 1008.

The acquisition unit 8001 in the device 8000 shown in FIG. 10 may be equivalent to the communication interface 9003 in the device 9000 shown in FIG. 11, and the corresponding training image may be obtained through the communication interface 9003. Alternatively, the acquisition unit 8001 may be equivalent to the processor 9002, and the training image may be obtained from the memory 9001 through the processor 9002, or the training image may be obtained from the outside through the communication interface 9003. In addition, the training unit 8002 in the device 8000 may be equivalent to the processor 9002 in the device 9000.

FIG11 is a schematic diagram of the hardware structure of a training device for a person re-identification network according to an embodiment of the present application. The training device 9000 for a person re-identification network shown in FIG11 (the device 9000 may be a computer device) includes a memory 9001, a processor 9002, a communication interface 9003, and a bus 9004. The memory 9001, the processor 9002, and the communication interface 9003 are connected to each other through the bus 9004.

The memory 9001 may be a read-only memory (ROM), a static storage device, a dynamic storage device or a random access memory (RAM). The memory 9001 may store a program. When the program stored in the memory 9001 is executed by the processor 9002, the processor 9002 is used to execute each step of the training method of the pedestrian re-identification network of the embodiment of the present application.

Processor 9002 can adopt a general-purpose central processing unit (CPU), a microprocessor, an application specific integrated circuit (ASIC), a graphics processing unit (GPU) or one or more integrated circuits to execute relevant programs to implement the training method of the pedestrian re-identification network of the embodiment of the present application.

The processor 9002 may also be an integrated circuit chip with signal processing capability. In the implementation process, each step of the training method of the pedestrian re-identification network of the present application may be completed by hardware integrated logic circuits in the processor 9002 or software instructions.

The processor 9002 may also be a general-purpose processor, a digital signal processor (DSP), an application-specific integrated circuit (ASIC), a field programmable gate array (FPGA) or other programmable logic devices, discrete gates or transistor logic devices, or discrete hardware components. The methods, steps, and logic block diagrams disclosed in the embodiments of the present application may be implemented or executed. The general-purpose processor may be a microprocessor or the processor may be any conventional processor, etc. The steps of the method disclosed in the embodiments of the present application may be directly embodied as being executed by a hardware decoding processor, or may be executed by a combination of hardware and software modules in the decoding processor. The software module may be located in a mature storage medium in the art, such as a random access memory, a flash memory, a read-only memory, a programmable read-only memory, or an electrically erasable programmable memory, a register, etc. The storage medium is located in the memory 9001, and the processor 9002 reads the information in the memory 9001, and completes the functions required to be performed by the units included in the training device of the pedestrian re-identification network in combination with its hardware, or executes the training method of the pedestrian re-identification network of the embodiment of the present application.

The communication interface 9003 uses a transceiver device such as, but not limited to, a transceiver to implement communication between the device 9000 and other devices or a communication network. For example, the image to be recognized can be obtained through the communication interface 9003.

The bus 9004 may include a path for transmitting information between various components of the device 9000 (eg, the memory 9001 , the processor 9002 , and the communication interface 9003 ).

FIG12 is a schematic block diagram of a pedestrian re-identification device according to an embodiment of the present application. The pedestrian re-identification device 10000 shown in FIG12 includes an acquisition unit 10001 and a recognition unit 10002 .

The acquisition unit 10001 and the recognition unit 10002 can be used to execute the pedestrian re-recognition method of the embodiment of the present application.

Specifically, the acquisition unit 10001 may execute the above step 6001, and the identification unit 10002 may execute the above step 6002.

The acquisition unit 10001 in the device 10000 shown in Figure 12 above can be equivalent to the communication interface 11003 in the device 11000 shown in Figure 13, and the image to be identified can be obtained through the communication interface 11003, or the acquisition unit 10001 can be equivalent to the processor 11002. At this time, the image to be identified can be obtained from the memory 11001 through the processor 11002, or the image to be identified can be obtained from the outside through the communication interface 11003.

In addition, the identification unit 10002 in the device 10000 shown in FIG. 12 is equivalent to the processor 11002 in the device 11000 shown in FIG. 13 .

FIG13 is a schematic diagram of the hardware structure of the pedestrian re-identification device of the embodiment of the present application. Similar to the above-mentioned device 10000, the pedestrian re-identification device 11000 shown in FIG13 includes a memory 11001, a processor 11002, a communication interface 11003 and a bus 11004. Among them, the memory 11001, the processor 11002, and the communication interface 11003 are connected to each other through the bus 11004.

The memory 11001 may be a ROM, a static storage device, or a RAM. The memory 11001 may store a program. When the program stored in the memory 11001 is executed by the processor 11002, the processor 11002 and the communication interface 11003 are used to execute the various steps of the pedestrian re-identification method of the embodiment of the present application.

Processor 11002 can be a general-purpose CPU, microprocessor, ASIC, GPU or one or more integrated circuits, which are used to execute relevant programs to implement the functions required to be performed by the units in the pedestrian re-identification device of the embodiment of the present application, or to execute the pedestrian re-identification method of the embodiment of the present application.

The processor 11002 may also be an integrated circuit chip with signal processing capability. In the implementation process, each step of the pedestrian re-identification method of the embodiment of the present application may be completed by an integrated logic circuit of hardware in the processor 11002 or by instructions in the form of software.

The processor 11002 may also be a general-purpose processor, DSP, ASIC, FPGA or other programmable logic device, discrete gate or transistor logic device, discrete hardware component. The methods, steps and logic block diagrams disclosed in the embodiments of the present application may be implemented or executed. The general-purpose processor may be a microprocessor or the processor may be any conventional processor, etc. The steps of the method disclosed in the embodiments of the present application may be directly embodied as being executed by a hardware decoding processor, or may be executed by a combination of hardware and software modules in the decoding processor. The software module may be located in a mature storage medium in the art such as a random access memory, a flash memory, a read-only memory, a programmable read-only memory or an electrically erasable programmable memory, a register, etc. The storage medium is located in the memory 11001, and the processor 11002 reads the information in the memory 11001, and combines its hardware to complete the functions required to be performed by the units included in the pedestrian re-identification device of the embodiment of the present application, or executes the pedestrian re-identification method of the embodiment of the present application.

The communication interface 11003 uses a transceiver device such as, but not limited to, a transceiver to implement communication between the device 11000 and other devices or a communication network. For example, the image to be recognized can be obtained through the communication interface 11003.

The bus 11004 may include a path for transmitting information between various components of the device 11000 (eg, the memory 11001 , the processor 11002 , and the communication interface 11003 ).

It should be noted that although the above-mentioned apparatus 9000 and apparatus 11000 only show a memory, a processor, and a communication interface, in the specific implementation process, those skilled in the art should understand that the apparatus 9000 and apparatus 11000 may also include other devices necessary for normal operation. At the same time, according to specific needs, those skilled in the art should understand that the apparatus 9000 and apparatus 11000 may also include hardware devices for implementing other additional functions. In addition, those skilled in the art should understand that the apparatus 9000 and apparatus 11000 may also only include the devices necessary for implementing the embodiments of the present application, and do not necessarily include all the devices shown in Figures 11 and 13.

Those of ordinary skill in the art will appreciate that the units and algorithm steps of each example 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 performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of this application.

Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working processes of the systems, devices and units described above can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

In the several embodiments provided in the present application, it should be understood that the disclosed systems, devices and methods can be implemented in other ways. For example, the device embodiments described above are only schematic. For example, the division of the units is only a logical function division. There may be other division methods in actual implementation, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of devices or units, which can be electrical, mechanical or other forms.

The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

In addition, each functional unit in each embodiment of the present application 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.

If the functions are implemented in the form of software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present application can be essentially or partly embodied in the form of a software product that contributes to the prior art. The computer software product is stored in a storage medium and includes several instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the methods described in each embodiment of the present application. The aforementioned storage medium includes: various media that can store program codes, such as a USB flash drive, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk.

The above is only a specific implementation of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art who is familiar with the present technical field can easily think of changes or substitutions within the technical scope disclosed in the present application, which should be included in the protection scope of the present application. Therefore, the protection scope of the present application should be based on the protection scope of the claims.

Claims (14)

1. A method for training a person re-identification network, comprising: Step 1: obtaining M training images and annotation data of the M training images, wherein each of the M training images includes a pedestrian, and the annotation data of each training image includes a bounding box where the pedestrian in each training image is located and pedestrian identification information, wherein different pedestrians correspond to different pedestrian identification information, and among the M training images, the training images with the same pedestrian identification information are from the same image capture device, and M is an integer greater than 1; Step 2: Initializing the network parameters of the pedestrian re-identification network to obtain initial values of the network parameters of the pedestrian re-identification network; Step 3: inputting a batch of training images from the M training images into the person re-identification network for feature extraction to obtain a feature vector of each training image in the batch of training images; The batch of training images includes N anchor images, the N anchor images are any N training images in the batch of training images, each of the N anchor images corresponds to a most difficult positive sample image, a first most difficult negative sample image and a second most difficult negative sample image, and N is a positive integer; The most difficult positive sample image corresponding to each anchor image is a training image in the batch of training images that has the same pedestrian identification information as each anchor image and has the farthest distance from the feature vector of each anchor image; the first most difficult negative sample image corresponding to each anchor image is a training image in the batch of training images that comes from the same image capture device as each anchor image, has different pedestrian identification information from each anchor image, and has the shortest distance from the feature vector of each anchor image; the second most difficult negative sample image corresponding to each anchor image is a training image in the batch of training images that comes from a different image capture device from each anchor image, has different pedestrian identification information from each anchor image, and has the shortest distance from the feature vector of each anchor image; Step 4: determining a function value of a loss function according to the feature vectors of the batch of training images, wherein the function value of the loss function is obtained by averaging the function values of N first loss functions; Wherein, the function value of each first loss function of the N first loss functions is calculated according to the first difference and the second difference corresponding to each anchor image of the N anchor images, the first difference corresponding to each anchor image is the difference between the most difficult positive sample distance corresponding to each anchor image and the second most difficult negative sample distance corresponding to each anchor image, the second difference corresponding to each anchor image is the difference between the second most difficult negative sample distance corresponding to each anchor image and the first most difficult negative sample distance corresponding to each anchor image, the most difficult positive sample distance corresponding to each anchor image is the distance between the feature vector of the most difficult positive sample image corresponding to each anchor image and the feature vector of each anchor image, the second most difficult negative sample distance corresponding to each anchor image is the distance between the feature vector of the second most difficult negative sample image corresponding to each anchor image and the feature vector of each anchor image, and the first most difficult negative sample distance corresponding to each anchor image is the distance between the feature vector of the first most difficult negative sample image corresponding to each anchor image and the feature vector of each anchor image; Step 5: updating the network parameters of the person re-identification network according to the function value of the loss function; Repeat steps 3 to 5 until the pedestrian re-identification network meets the preset requirements. 2. The training method according to claim 1, wherein the person re-identification network meets preset requirements, including: The pedestrian re-identification network meets the preset requirements when at least one of the following conditions is met: The number of training times of the pedestrian re-identification network is greater than or equal to a preset number of times; The function value of the loss function is less than or equal to a preset threshold; The recognition performance of the pedestrian re-identification network meets the preset requirements. 3. The training method according to claim 2, wherein the function value of the loss function is less than or equal to a preset threshold, comprising: The first difference is smaller than a first preset threshold, and the second difference is smaller than a second preset threshold. 4. The training method according to any one of claims 1 to 3, characterized in that the M training images are training images from multiple image capturing devices, wherein the labeling data of the training images from different image capturing devices are obtained by separate labeling. 5. A pedestrian re-identification method, comprising: Obtain an image to be recognized; Processing the image to be identified using a person re-identification network to obtain a feature vector of the image to be identified, wherein the person re-identification network is trained according to the training method according to any one of claims 1 to 4; The recognition result of the image to be recognized is obtained by comparing the feature vector of the image to be recognized with the feature vector of the existing pedestrian image. 6. A training device for a person re-identification network, comprising: An acquisition unit, used to execute step 1; Step 1: obtaining M training images and annotation data of the M training images, wherein each of the M training images includes a pedestrian, and the annotation data of each training image includes a bounding box where the pedestrian in each training image is located and pedestrian identification information, wherein different pedestrians correspond to different pedestrian identification information, and among the M training images, the training images with the same pedestrian identification information are from the same image capture device, and M is an integer greater than 1; A training unit, used to execute step 2; Step 2: Initializing the network parameters of the pedestrian re-identification network to obtain initial values of the network parameters of the pedestrian re-identification network; The training unit is further used to repeatedly execute steps 3 to 5 until the pedestrian re-identification network meets preset requirements; Step 3: inputting a batch of training images from the M training images into the person re-identification network for feature extraction to obtain a feature vector of each training image in the batch of training images; The batch of training images includes N anchor images, the N anchor images are any N training images in the batch of training images, each of the N anchor images corresponds to a most difficult positive sample image, a first most difficult negative sample image and a second most difficult negative sample image, and N is a positive integer; The most difficult positive sample image corresponding to each anchor image is a training image in the batch of training images that has the same pedestrian identification information as each anchor image and has the farthest distance from the feature vector of each anchor image; the first most difficult negative sample image corresponding to each anchor image is a training image in the batch of training images that comes from the same image capture device as each anchor image, has different pedestrian identification information from each anchor image, and has the shortest distance from the feature vector of each anchor image; the second most difficult negative sample image corresponding to each anchor image is a training image in the batch of training images that comes from a different image capture device from each anchor image, has different pedestrian identification information from each anchor image, and has the shortest distance from the feature vector of each anchor image; Step 4: determining a function value of a loss function according to the feature vectors of the batch of training images, wherein the function value of the loss function is obtained by averaging the function values of N first loss functions; Wherein, the function value of each first loss function of the N first loss functions is calculated according to the first difference and the second difference corresponding to each anchor image of the N anchor images, the first difference corresponding to each anchor image is the difference between the most difficult positive sample distance corresponding to each anchor image and the second most difficult negative sample distance corresponding to each anchor image, the second difference corresponding to each anchor image is the difference between the second most difficult negative sample distance corresponding to each anchor image and the first most difficult negative sample distance corresponding to each anchor image, the most difficult positive sample distance corresponding to each anchor image is the distance between the feature vector of the most difficult positive sample image corresponding to each anchor image and the feature vector of each anchor image, the second most difficult negative sample distance corresponding to each anchor image is the distance between the feature vector of the second most difficult negative sample image corresponding to each anchor image and the feature vector of each anchor image, and the first most difficult negative sample distance corresponding to each anchor image is the distance between the feature vector of the first most difficult negative sample image corresponding to each anchor image and the feature vector of each anchor image; Step 5: Update the network parameters of the person re-identification network according to the function value of the loss function. 7. The training device according to claim 6, wherein the pedestrian re-identification network meets the preset requirements, including: The pedestrian re-identification network meets the preset requirements when at least one of the following conditions is met: The number of training times of the pedestrian re-identification network is greater than or equal to a preset number of times; The function value of the loss function is less than or equal to a preset threshold; The recognition performance of the pedestrian re-identification network meets the preset requirements. 8. The training device according to claim 7, wherein the function value of the loss function is less than or equal to a preset threshold, comprising: The first difference is smaller than a first preset threshold, and the second difference is smaller than a second preset threshold. 9. The training device as described in any one of claims 6-8 is characterized in that the M training images are training images from multiple image capturing devices, wherein the labeling data of the training images from different image capturing devices are obtained by separate labeling. 10. A pedestrian re-identification device, comprising: An acquisition unit, used for acquiring an image to be recognized; A recognition unit, configured to process the image to be recognized by using a person re-recognition network to obtain a feature vector of the image to be recognized, wherein the person re-recognition network is trained according to the training method according to any one of claims 1 to 4; The recognition unit is further used to compare the feature vector of the image to be recognized with the feature vector of the existing pedestrian image to obtain the recognition result of the image to be recognized. 11. A computer-readable storage medium, characterized in that the computer-readable medium stores program codes for execution by a device, wherein the program codes include codes for executing the training method according to any one of claims 1 to 4. 12. A computer-readable storage medium, characterized in that the computer-readable medium stores program codes for execution by a device, the program codes including codes for executing the pedestrian re-identification method as claimed in claim 5. 13. A chip, characterized in that the chip comprises a processor and a data interface, and the processor reads instructions stored in a memory through the data interface to execute the training method as described in any one of claims 1-4. 14. A chip, characterized in that the chip comprises a processor and a data interface, and the processor reads instructions stored in a memory through the data interface to execute the pedestrian re-identification method as claimed in claim 5.