Domain adaptive target detection method based on big and small model feature collaboration

Abstract

The invention relates to the technical field of target detection, and particularly provides a domain adaptive target detection method based on large and small model feature collaboration. The method comprises the following steps: establishing a global local enhancement network; according to the global and local enhancement network, building a large and small model collaborative feature interaction enhancement network; connecting the global local enhancement network and the feature interaction enhancement network with a detection head, and constructing a domain adaptive target detection model; according to the method, the domain self-adaptive target detection model is trained, the trained domain self-adaptive target detection model is packaged and deployed to realize target identification, the domain self-adaptive ability of the model in different scenes is improved, and more efficient target identification and positioning are realized by effectively reducing the distribution difference between the source domain and the target domain.

Description

Technical Field

[0001] The present invention relates to the technical field of target detection, and in particular to a domain adaptive target detection method based on large and small model feature collaboration. Background Art

[0002] Domain adaptation is an important transfer learning strategy that aims to solve the problem of inconsistent distribution of training and test data. With the rapid development of deep learning technology, domain adaptation is increasingly used in computer vision, especially in object detection tasks. Object detection usually relies on a large amount of labeled data for model training, but in many practical scenarios, it is often very difficult and costly to obtain high-quality labeled data.

[0003] Domain adaptation technology improves detection performance in the target domain (unlabeled data) by leveraging knowledge in the source domain (labeled data). The core of this method is to reduce the distribution difference between the source domain and the target domain to achieve better migration effect. Traditional domain adaptation methods mainly focus on feature-level alignment, which promotes the generalization ability of the model in new domains by minimizing the distance between the feature distribution of the source domain and the target domain.

[0004] However, in practical applications, domain adaptation still faces some challenges. First, there may be significant differences in target morphology, background complexity, and environmental conditions between different domains, which makes simple feature alignment methods limited in effectiveness. Second, the size and scale of the target also affect the performance of the detection model. Objects of different sizes may have different performances in the same image, so it is necessary to consider the fusion and utilization of multi-scale features. Summary of the invention

[0005] In view of this, the present invention provides a domain adaptive target detection method based on large and small model feature collaboration, so as to improve the domain adaptation ability of the model in different scenarios, and achieve more efficient target recognition and positioning by effectively reducing the distribution difference between the source domain and the target domain.

[0006] In a first aspect, the present invention provides a domain adaptive object detection method based on large and small model feature collaboration, the method comprising:

[0007] Step S1, building a global local enhancement network;

[0008] Step S2: Building a feature interaction enhancement network for large and small model collaboration based on the global and local enhancement network in step S1;

[0009] Step S3, connecting the global local enhancement network in step S1 and the feature interaction enhancement network in step S2 to the detection head to build a domain adaptive object detection model;

[0010] Step S4: training the domain adaptive target detection model constructed in step S3, and encapsulating and deploying the trained domain adaptive target detection model to achieve target recognition.

[0011] Optionally, the step S1 includes:

[0012] The global-local enhancement network consists of a global enhancement module and a local enhancement module; the global enhancement module and the local enhancement module enhance the image in parallel;

[0013] The global enhancement module includes color space conversion, initial irradiance estimation, fractional differential mask and color space inverse conversion; the local enhancement module includes random target border extraction and multi-scale retinal enhancement with color restoration.

[0014] Optionally, the global enhancement module comprises:

[0015] a. Convert the source domain image to be enhanced from RGB color space to YCbCr color space, and extract the converted brightness component as the initial image I′(x, y). The formula for generating the initial image is:

[0016] I'(x,y)=0.299R(x,y)+0.587G(x,y)+0.114B(x,y);

[0017] b. Use the Retinex algorithm based on fast guided filtering to process the initial image I′(x, y) and estimate the initial illumination component L′(x, y). According to the Retinex theory, the reflection component R′(x, y) can be calculated by the following formula:

[0018]

[0019] Where GF represents the fast guided filter function, r is the control parameter of the filter radius, is a parameter for controlling the filtering strength, τ represents a parameter for adjusting the edge preservation characteristics, ω represents a parameter for controlling the filter response, and δ k Represents the weight coefficient of each calculation, d x and d y represents the displacement in the x and y directions, ∈ k represents the disturbance term that changes with k, which is used to avoid the situation where the denominator is zero;

[0020] c. Use adaptive fractional differential mask to process the reflection component R′(x, y) to obtain the enhanced reflection component R″(x, y). The generation formula is:

[0021] R″(x,y)=R′(x,y)×M(x,y,α);

[0022]

[0023] Where M(x, y, α) represents a fractional differential mask with fractional order α; R″(xk, yk) represents the pixel value of the local area of ​​the image, which is used to perform the convolution operation;

[0024] d. Combine the enhanced reflection component R″(x,y) with the original color difference components Cb(x,y) and Cr(x,y), and convert the YCbCr color space back to the RGB color space to obtain the final global enhanced image I g (x,y).

[0025] Optionally, the local enhancement module comprises:

[0026] e. Randomly set a pixel coordinate as the center point within the provided real border and generate a horizontal or vertical border to be enhanced around it;

[0027] f. Image I in the border area to be enhanced b (x,y) is enhanced by a multi-scale retinal algorithm to obtain the final local enhanced image I l (x,y), the generating formula is:

[0028]

[0029] Among them, ω n is the weight of small-scale, medium-scale, and large-scale brightness estimation; G n (x,y) is the Gaussian filter function;

[0030] g. The image finally obtained by the global local enhancement network is expressed as:

[0031] I a =βI g +(1-β)I l ;

[0032] Where β represents the fusion factor.

[0033] Optionally, step S2 includes:

[0034] The feature interaction enhancement network consists of a feature interaction module and a feature causal attention alignment module.

[0035] Optionally, the feature interaction module comprises:

[0036] h. Extract features from the encoder of the segmentation model and the encoder of Faster RCNN respectively, and extract the features F of the segmentation model. S Through MLP and basic features FR To align the dimensions, the expression is:

[0037]

[0038] Among them, F′ S represents the auxiliary features after MLP alignment, ensuring that they are consistent with the basic features F R The number of channels is consistent;

[0039] i. Align the auxiliary feature F′ S and the basic feature F R The self-attention mechanism is applied respectively, and its expression is:

[0040] F″ S =SA(F′ S ),F″ R =SA(F R );

[0041] j. Introduce the cross attention mechanism to add auxiliary features F″ S and basic feature F″ R To interact, the expression is:

[0042]

[0043] Where Q, K, and V represent query, key, and value matrices, respectively, and d k Indicates the dimension of the key, used to scale the result of the dot product;

[0044] k, the auxiliary feature F″ after self-attention processing S 、Basic feature F″ R and cross-attention fusion feature F C Aggregate to generate the final interaction feature, which is expressed as:

[0045]

[0046] where ω i 、v i , γ i represents the non-negative weight of each feature.

[0047] Optionally, the feature causal attention alignment module comprises:

[0048] m. Extract feature maps through the Faster RCNN feature extraction network and the segmentation model feature extraction network respectively; based on these feature maps, calculate the corresponding attention map A R (I a ) and A S (I a ), whose expression is:

[0049] A R (I a )=σ(f R (I a )),A S (I a )=σ(f S (I a ));

[0050] where f R ,f S They represent the faster RCNN and segmentation model feature extraction networks respectively, and σ represents the Sigmoid function, which is used to normalize the feature map to the range of [0,1];

[0051] n. The attention map is processed by the binarization function B(·), and the Dice loss is used to measure the similarity between them, which is expressed as:

[0052]

[0053] Where B(·) represents the pixel-level binarization operation.

[0054] Optionally, step S3 includes:

[0055] The detection head is sequentially connected to a global enhancement module, a local enhancement module, a feature interaction module and a feature causal attention alignment module, and after the connection, a domain adaptive object detection model is constructed.

[0056] Optionally, step S4 includes:

[0057] v. The domain adaptive target detection model adopts a teacher-student network and updates the weights through exponential moving average, and its expression is:

[0058] θ t (t) = αθ t (t-1)+(1-α)θ s (t);

[0059] w. During training, the supervised loss supLoss of the student network comes from the region proposal network and the ROI pooling detection head; when pre-training is completed, the weights of the student network are transferred to the teacher network; then, the unsupervised loss usupLoss is guided by the matching pseudo-labels generated by the teacher network, and its expression is:

[0060]

[0061] where f s (·) represents the detector of the student network, f s(·) denotes the detector of the teacher network;

[0062] s, the total loss function TotalLoss in training, its expression is:

[0063] TotalLoss=supLoss+μ1usupLoss+μ2CLoss;

[0064] t. Input the training set, validation set, and test set into the constructed target detection model and train it. Select the target detection model that has been trained for 90K rounds for TensorRT acceleration and encapsulate it into a DLL file.

[0065] In the technical solution provided by the present invention, the method includes building a global local enhancement network; building a feature interaction enhancement network for collaborative large and small models based on the global local enhancement network; connecting the global local enhancement network and the feature interaction enhancement network to a detection head to construct a domain adaptive target detection model; training the domain adaptive target detection model, and encapsulating and deploying the trained domain adaptive target detection model to achieve target recognition. The method improves the domain adaptation capability of the model in different scenarios, and achieves more efficient target recognition and positioning by effectively reducing the distribution difference between the source domain and the target domain. BRIEF DESCRIPTION OF THE DRAWINGS

[0066] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0067] Figure 1 A flowchart of a domain adaptive object detection method provided by an embodiment of the present invention;

[0068] Figure 2 A schematic diagram of a global local enhancement network provided by an embodiment of the present invention;

[0069] Figure 3 A schematic diagram of a domain adaptive object detection model provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0070] In order to make the purpose, technical solution and advantages of the embodiments of the present invention clearer, the technical solution in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0071] It should be clear that the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0072] The terms used in the embodiments of the present invention are only for the purpose of describing specific embodiments, and are not intended to limit the present invention. The singular forms "a", "said" and "the" used in the embodiments of the present invention are also intended to include plural forms, unless the context clearly indicates other meanings.

[0073] It should be understood that the term "and / or" used in this article is only a description of the association relationship of associated objects, indicating that there can be three relationships. For example, A and / or B can represent: A exists alone, A and B exist at the same time, and B exists alone. In addition, the character " / " in this article generally indicates that the associated objects before and after are in an "or" relationship.

[0074] The word "if" as used herein may be interpreted as "at the time of" or "when" or "in response to determining" or "in response to detecting", depending on the context. Similarly, the phrases "if it is determined" or "if (stated condition or event) is detected" may be interpreted as "when it is determined" or "in response to determining" or "when detecting (stated condition or event)" or "in response to detecting (stated condition or event)", depending on the context.

[0075] Figure 1 A flowchart of a domain adaptive target detection method provided by an embodiment of the present invention is shown in FIG. Figure 1 As shown, the method includes:

[0076] Step S1: Build a global and local enhancement network.

[0077] In the embodiment of the present invention, Figure 2 As shown, step S1 includes:

[0078] The global-local enhancement network consists of a global enhancement module and a local enhancement module; the global enhancement module and the local enhancement module enhance the image in parallel;

[0079] The global enhancement module includes color space conversion, initial irradiance estimation, fractional differential mask and color space inverse conversion; the local enhancement module includes random target border extraction and multi-scale retinal enhancement with color restoration.

[0080] In the embodiment of the present invention, the global enhancement module comprises:

[0081] a. Convert the source domain image to be enhanced from RGB color space to YCbCr color space, and extract the converted brightness component as the initial image I′(x, y). The formula for generating the initial image is:

[0082] I'(x,y)=0.299R(x,y)+0.587G(x,y)+0.114B(x,y);

[0083] b. Use the Retinex algorithm based on fast guided filtering to process the initial image I′(x, y) and estimate the initial illumination component L′(x, y). According to the Retinex theory, the reflection component R′(x, y) can be calculated by the following formula:

[0084]

[0085] Where GF represents the fast guided filter function, r is the control parameter of the filter radius, is a parameter for controlling the filtering strength, τ represents a parameter for adjusting the edge preservation characteristics, ω represents a parameter for controlling the filter response, and δ k Represents the weight coefficient of each calculation, d x and d y represents the displacement in the x and y directions, ∈ k represents the disturbance term that changes with k, which is used to avoid the situation where the denominator is zero;

[0086] c. Use adaptive fractional differential mask to process the reflection component R′(x, y) to obtain the enhanced reflection component R″(x, y). The generation formula is:

[0087] R″(x,y)=R′(x,y)×M(x,y,α);

[0088]

[0089] Where M(x, y, α) represents a fractional differential mask with fractional order α; R″(xk, yk) represents the pixel value of the local area of ​​the image, which is used to perform the convolution operation;

[0090] d. Combine the enhanced reflection component R″(x,y) with the original color difference components Cb(x,y) and Cr(x,y), and convert the YCbCr color space back to the RGB color space to obtain the final global enhanced image I g (x,y).

[0091] In the embodiment of the present invention, the local enhancement module comprises:

[0092] e. Randomly set a pixel coordinate as the center point within the provided real border and generate a horizontal or vertical border to be enhanced around it;

[0093] f. Image I in the border area to be enhanced b (x,y) is enhanced by a multi-scale retinal algorithm to obtain the final local enhanced image I l (x,y), the generating formula is:

[0094]

[0095] Among them, ω n is the weight of small-scale, medium-scale, and large-scale brightness estimation; G n (x,y) is the Gaussian filter function;

[0096] g. The image finally obtained by the global local enhancement network is expressed as:

[0097] I a =βI g +(1-β)I l ;

[0098] Where β represents the fusion factor.

[0099] Step S2: Based on the global-local enhancement network in step S1, a feature interaction enhancement network for collaborative large and small models is constructed.

[0100] In the embodiment of the present invention, Figure 3 As shown, step S2 includes:

[0101] The feature interaction enhancement network consists of a feature interaction module and a feature causal attention alignment module.

[0102] In the embodiment of the present invention, the feature interaction module comprises:

[0103] h. Extract features from the encoder of the segmentation model and the encoder of the Faster RCNN, i.e., the 50th layer of the residual network, respectively, and extract the features of the segmentation model F S Through the multi-layer perceptron MLP and the basic feature F R To align the dimensions, the expression is:

[0104]

[0105] Among them, F′ S represents the auxiliary features after MLP alignment, ensuring that they are consistent with the basic features F R The number of channels is consistent;

[0106] i. Align the auxiliary feature F′ Sand the basic feature F R The self-attention mechanism is applied respectively, and its expression is:

[0107] F″ S =SA(F′ S ),F″ R =SA(F R );

[0108] j. Introduce the cross attention mechanism to add auxiliary features F″ S and basic feature F″ R To interact, the expression is:

[0109]

[0110] Where Q, K, and V represent query, key, and value matrices, respectively, and d k Indicates the dimension of the key, used to scale the result of the dot product;

[0111] k, the auxiliary feature F″ after self-attention processing S 、Basic feature F″ R and cross-attention fusion feature F C Aggregate to generate the final interaction feature, which is expressed as:

[0112]

[0113] where ω i 、v i , γ i represents the non-negative weight of each feature.

[0114] In the embodiment of the present invention, the feature causal attention alignment module comprises:

[0115] m. Extract feature maps through the Faster RCNN feature extraction network and the segmentation model feature extraction network respectively; based on these feature maps, calculate the corresponding attention map A R (I a ) and A S (I a ), whose expression is:

[0116] A R (I a )=σ(f R (I a )0,A S (I a )=σ(f S (I a ));

[0117] where f R ,fS They represent the faster RCNN and segmentation model feature extraction networks respectively, σ represents the Sigmoid function, which is used to normalize the feature map to the range of [0,1];

[0118] n. The attention map is processed by the binarization function B(·), and the Dice loss is used to measure the similarity between them, which is expressed as:

[0119]

[0120] Where B(·) represents the pixel-level binarization operation.

[0121] Step S3: Connect the global local enhancement network in step S1 and the feature interaction enhancement network in step S2 to the detection head to build a domain adaptive object detection model.

[0122] In the embodiment of the present invention, step S3 includes:

[0123] The detection head is sequentially connected to a global enhancement module, a local enhancement module, a feature interaction module and a feature causal attention alignment module, and after the connection, a domain adaptive object detection model is constructed.

[0124] Step S4: training the domain adaptive target detection model constructed in step S3, and encapsulating and deploying the trained domain adaptive target detection model to achieve target recognition.

[0125] In the embodiment of the present invention, step S4 includes:

[0126] v. The domain adaptive target detection model adopts a teacher-student network and updates the weights through exponential moving average, and its expression is:

[0127] θ t (t) = αθ t (t-1)+(1-α)θ s (t);

[0128] w. During training, the supervised loss supLoss of the student network comes from the region proposal network and the ROI pooling detection head; when pre-training is completed, the weights of the student network are transferred to the teacher network; then, the unsupervised loss usupLoss is guided by the matching pseudo-labels generated by the teacher network, and its expression is:

[0129]

[0130] where f s (·) represents the detector of the student network, f s (·) denotes the detector of the teacher network;

[0131] s, the total loss function TotalLoss in training, its expression is:

[0132] TotalLoss=supLoss+μ1usupLoss+μ2CLoss;

[0133] t. Input the training set, validation set, and test set into the constructed target detection model and train it. Select the target detection model that has been trained for 90K rounds for TensorRT acceleration and encapsulate it into a DLL file.

[0134] This invention combines the features of different models to enhance the adaptability of the basic model to various targets, while improving the detection performance of the target domain. This method can not only effectively reduce the feature differences between the source domain and the target domain, but also achieve better detection results in multiple scenarios, providing a new idea and path for the development of domain adaptive target detection technology.

[0135] In the technical solution provided by the present invention, the method includes building a global local enhancement network; building a feature interaction enhancement network for collaborative large and small models based on the global local enhancement network; connecting the global local enhancement network and the feature interaction enhancement network to a detection head to construct a domain adaptive target detection model; training the domain adaptive target detection model, and encapsulating and deploying the trained domain adaptive target detection model to achieve target recognition. The method improves the domain adaptation capability of the model in different scenarios, and achieves more efficient target recognition and positioning by effectively reducing the distribution difference between the source domain and the target domain.

[0136] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the scope of protection of the present invention.

Claims

1. A domain adaptive object detection method based on large and small model feature collaboration, characterized in that: The method comprises: Step S1, building a global local enhancement network; Step S2: Building a feature interaction enhancement network for large and small model collaboration based on the global and local enhancement network in step S1; Step S3, connecting the global local enhancement network in step S1 and the feature interaction enhancement network in step S2 to the detection head to build a domain adaptive object detection model; Step S4: training the domain adaptive target detection model constructed in step S3, and encapsulating and deploying the trained domain adaptive target detection model to achieve target recognition.

2. The method according to claim 1, characterized in that The step S1 comprises: The global-local enhancement network consists of a global enhancement module and a local enhancement module; the global enhancement module and the local enhancement module enhance the image in parallel; The global enhancement module includes color space conversion, initial irradiance estimation, fractional differential mask and color space inverse conversion; the local enhancement module includes random target border extraction and multi-scale retinal enhancement with color restoration.

3. The method according to claim 2, characterized in that The global enhancement module comprises: a. Convert the source domain image to be enhanced from RGB color space to YCbCr color space, and extract the converted brightness component as the initial image I′(x, y). The formula for generating the initial image is: I′(x,y)=0.299R(x,y)+0.587G(x,y)+0.114B(x,y); b. Use the Retinex algorithm based on fast guided filtering to process the initial image I′(x, y) and estimate the initial illumination component L′(x, y). According to the Retinex theory, the reflection component R′(x, y) can be calculated by the following formula: L′(x,y)=GF(I′(x,y),r,θ,τ,ω); Where GF represents the fast guided filter function, r is the control parameter of the filter radius, θ is the parameter for controlling the filter strength, τ is the parameter for adjusting the edge preservation characteristics, ω is the parameter for controlling the response of the filter, and δ k Represents the weight coefficient of each calculation, d x and d y represents the displacement in the x and y directions, ∈ k represents the disturbance term that changes with k, which is used to avoid the situation where the denominator is zero; c. Use adaptive fractional differential mask to process the reflection component R′(x, y) to obtain the enhanced reflection component R″(x, y). The generation formula is: R″(x,y)=R′(x,y)×M(x,y,α); Where M(x, y, α) represents a fractional differential mask with fractional order α; R″(xk, yk) represents the pixel value of the local area of ​​the image, which is used to perform the convolution operation; d. Combine the enhanced reflection component R″(x,y) with the original color difference components Cb(x,y) and Cr(x,y), and convert the YCbCr color space back to the RGB color space to obtain the final global enhanced image I g (x,y).

4. The method according to claim 2, characterized in that: The local enhancement module comprises: e. Randomly set a pixel coordinate as the center point within the provided real border and generate a horizontal or vertical border to be enhanced around it; f. Image I in the border area to be enhanced b (x,y) is enhanced by a multi-scale retinal algorithm to obtain the final local enhanced image I l (x,y), the generating formula is: Among them, ω n is the weight of small-scale, medium-scale, and large-scale brightness estimation; G n (x,y) is the Gaussian filter function; g. The image finally obtained by the global local enhancement network is expressed as: I a =βI g +(1−β)I l 4 Where β represents the fusion factor.

5. The method according to claim 1, characterized in that The step S2 comprises: The feature interaction enhancement network consists of a feature interaction module and a feature causal attention alignment module.

6. The method according to claim 5, characterized in that The feature interaction module comprises: h. Extract features from the encoder of the segmentation model and the encoder of Faster RCNN respectively, and extract the features F of the segmentation model. S Through MLP and basic features F R To align the dimensions, the expression is: Among them, F′ S represents the auxiliary features after MLP alignment, ensuring that they are consistent with the basic features F R The number of channels is consistent; i. Align the auxiliary feature F′ S and the basic feature F R The self-attention mechanism is applied respectively, and its expression is: F″ S =SA(F′ S ),F″ R =SA(F R ); j. Introduce the cross attention mechanism to add auxiliary features F″ S and basic feature F″ R To interact, the expression is: Where Q, K, and V represent query, key, and value matrices, respectively, and d k Indicates the dimension of the key, used to scale the result of the dot product; k, the auxiliary feature F″ after self-attention processing S 、Basic feature F″ R and cross-attention fusion feature F C Aggregate to generate the final interaction feature, which is expressed as: where ω i 、v i , γ i represents the non-negative weight of each feature.

7. The method according to claim 5, characterized in that The feature causal attention alignment module comprises: m. Extract feature maps through the Faster RCNN feature extraction network and the segmentation model feature extraction network respectively; based on these feature maps, calculate the corresponding attention map A R (I a ) and A S (I a ), whose expression is: A R (I a )=σ(f R (I a )),A S (I a )=σ(f S (I a )); where f R ,f S They represent the faster RCNN and segmentation model feature extraction networks respectively, σ represents the Sigmoid function, which is used to normalize the feature map to the range of [0,1]; n. The attention map is processed by the binarization function B(·), and the Dice loss is used to measure the similarity between them, which is expressed as: Where B(·) represents the pixel-level binarization operation.

8. The method according to claim 1, characterized in that The step S3 comprises: The detection head is sequentially connected to a global enhancement module, a local enhancement module, a feature interaction module and a feature causal attention alignment module, and after the connection, a domain adaptive object detection model is constructed.

9. The method according to claim 1, characterized in that: The step S4 comprises: v. The domain adaptive target detection model adopts a teacher-student network and updates the weights through exponential moving average, and its expression is: i t (t)=αθ t (t-1)+(1-α)θ s (t); w. During training, the supervised loss supLoss of the student network comes from the region proposal network and the ROI pooling detection head; when pre-training is completed, the weights of the student network are transferred to the teacher network; then, the unsupervised loss usupLoss is guided by the matching pseudo-labels generated by the teacher network, and its expression is: where f s (·) represents the detector of the student network, f s (·) denotes the detector of the teacher network; s, the total loss function TotalLoss in training, its expression is: TotalLoss=supLoss+μ1usupLoss+μ2CLoss; t. Input the training set, validation set, and test set into the constructed target detection model and train it. Select the target detection model that has been trained for 90K rounds for TensorRT acceleration and encapsulate it into a DLL file.

Images