CN113902974A - Air combat threat target identification method based on convolutional neural network - Google Patents
Air combat threat target identification method based on convolutional neural network Download PDFInfo
- Publication number
- CN113902974A CN113902974A CN202111134920.9A CN202111134920A CN113902974A CN 113902974 A CN113902974 A CN 113902974A CN 202111134920 A CN202111134920 A CN 202111134920A CN 113902974 A CN113902974 A CN 113902974A
- Authority
- CN
- China
- Prior art keywords
- threat
- threat target
- air
- air combat
- layer
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
- 238000000034 method Methods 0.000 title claims abstract description 28
- 238000013527 convolutional neural network Methods 0.000 title claims abstract description 10
- 230000007246 mechanism Effects 0.000 claims abstract description 22
- 238000013507 mapping Methods 0.000 claims abstract description 10
- 238000010586 diagram Methods 0.000 claims description 15
- 238000011176 pooling Methods 0.000 claims description 14
- 239000013598 vector Substances 0.000 claims description 9
- 238000012549 training Methods 0.000 claims description 5
- 238000001514 detection method Methods 0.000 abstract description 6
- 230000008447 perception Effects 0.000 abstract description 5
- 230000005540 biological transmission Effects 0.000 abstract description 3
- 230000002708 enhancing effect Effects 0.000 abstract description 2
- 238000003062 neural network model Methods 0.000 description 14
- 230000007123 defense Effects 0.000 description 10
- 230000006870 function Effects 0.000 description 9
- 238000000605 extraction Methods 0.000 description 8
- 238000013528 artificial neural network Methods 0.000 description 7
- 238000004422 calculation algorithm Methods 0.000 description 7
- 230000008569 process Effects 0.000 description 4
- 238000011160 research Methods 0.000 description 4
- 238000004364 calculation method Methods 0.000 description 3
- 238000013135 deep learning Methods 0.000 description 3
- 230000002349 favourable effect Effects 0.000 description 3
- 206010063385 Intellectualisation Diseases 0.000 description 2
- 206010033307 Overweight Diseases 0.000 description 2
- 230000006872 improvement Effects 0.000 description 2
- 230000035899 viability Effects 0.000 description 2
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000007796 conventional method Methods 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000003909 pattern recognition Methods 0.000 description 1
- 238000004088 simulation Methods 0.000 description 1
- 238000012546 transfer Methods 0.000 description 1
- 238000012795 verification Methods 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/045—Combinations of networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/048—Activation functions
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
Abstract
The invention discloses an air combat threat target identification method based on a convolutional neural network, which comprises the steps of firstly adding an attention mechanism into a backbone network of a ResNet model, enhancing the characteristic expression capability of an interested region, reducing the influence of useless information on an identification result, and improving the region focusing capability of a detection model; and then, at the end of the ResNet model, the structure of a single full connection layer is changed into the structure of double full connection layers, and the threat type and the guidance type of the air combat threat target are respectively subjected to feature mapping so as to improve the recognition capability of the model and accelerate the convergence of the model. Through the two methods, the improved ResNet model is obtained, the type of the air combat threat target can be identified by using the improved ResNet model, and an air combat threat target identification task is realized. The method fully utilizes the channel attention and the space attention mechanism to guide the transmission of important characteristic information, and effectively improves the perception capability of the recognition model on the global information and the recognition accuracy of the air war threat target type.
Description
Technical Field
The invention belongs to the technical field of pattern recognition, and particularly relates to an air combat threat target recognition method.
Background
The difficult point problem to be solved urgently is how to effectively improve the terminal confrontation efficiency of the fighter and improve the terminal defense viability of the fighter in the modern air battle in the aspect of continuously enhanced air battle threat targets (such as missiles) because the attack and defense confrontation of the fighter and the air battle threat targets is intensified day by day. The air combat threat target type identification is an important component of airborne terminal threat perception, and various attribute characteristic information of the air combat threat target obtained by a plurality of information sources is fully utilized and combined reasoning is carried out according to relevant criteria to obtain accurate and reliable type estimation. The accurate identification of the type of the air combat threat target is the basis of defense decision making of the fighter, and the identification speed and precision have important influence on the generation of a defense strategy and the effect of defense measures. Countries such as america and russian have long been accumulating in the technical field of identification of air war threat object types, and a great deal of experience has been accumulated through years of research. But the method is strictly confidential to the related technology, and a complete and effective identification method or identification system is difficult to find in foreign published documents. The research of the related technology for identifying the type of the air war threat target in China is late, the air war threat target type identification in China scientific research institutions has the characteristics of noise and certain randomness of the acquisition sequence, the identification of the type of the airborne air war threat target is often regarded as a dynamic uncertainty reasoning process, the credibility of the entropy gain-based description evidence is constructed and is integrated into the Bayesian network reasoning process, and certain results are obtained. However, under the complex conditions of external interference and multiple targets, in the face of the requirement of high confidence level in the complex environment, the Bayesian network inference model is poor in learning capability and sensitive to the form of input data, so that the identification precision of the air war threat target type is low, and the requirement is difficult to meet. Under complex conditions, the technology for identifying the type of the threat target of the air combat in a long distance and high precision needs to be broken through urgently.
In recent years, deep learning research is continuously concerned by scholars at home and abroad, and along with the continuous upgrading and updating of air war threat targets (such as air-to-air/surface-to-air missiles), the intellectualization and anti-interference capability of the air war threat targets are continuously enhanced, so that the tail-end confrontation environment is increasingly complicated, and great challenge is formed on the tail-end confrontation viability of a fighter. The problems of incomplete sensor observation information, low recognition confidence coefficient and the like in the airborne terminal countermeasure process are solved by using an advanced intelligent technology, an intelligent fighter terminal defense perception decision integrated technology is developed, the terminal countermeasure defense effectiveness is improved, and a necessary trend is achieved, so that a target recognition method for multi-source sensor information needs to be developed by combining a neural network algorithm.
Disclosure of Invention
In order to overcome the defects of the prior art, the invention provides an air combat threat target identification method based on a convolutional neural network, which comprises the steps of firstly adding an attention mechanism into a backbone network of a ResNet model, enhancing the characteristic expression capability of an interested region, reducing the influence of useless information on an identification result, and improving the region focusing capability of a detection model; and then, at the end of the ResNet model, the structure of a single full connection layer is changed into the structure of double full connection layers, and the threat type and the guidance type of the air combat threat target are respectively subjected to feature mapping so as to improve the recognition capability of the model and accelerate the convergence of the model. Through the two methods, the improved ResNet model is obtained, the type of the air combat threat target can be identified by using the improved ResNet model, and an air combat threat target identification task is realized. The method fully utilizes the channel attention and the space attention mechanism to guide the transmission of important characteristic information, and effectively improves the perception capability of the recognition model on the global information and the recognition accuracy of the air war threat target type.
The technical scheme adopted by the invention for solving the technical problem comprises the following steps:
step 1: acquiring air combat threat target data to be identified, and constructing an air combat threat target data set;
step 2: assigning two labels to each air combat threat target data in the air combat threat target data set, wherein the labels are respectively a threat type label and a guidance type label of the air combat threat target, and the labels are adopted to distinguish different types of air combat threat targets;
and step 3: carrying out data link on N data of the same air combat threat target in the air combat threat target data set at different times to obtain an enhanced air combat threat target data set;
and 4, step 4: constructing an improved ResNet model;
step 4-1: attention modules are added after the 7 th layer, the 15 th layer, the 27 th layer and the 33 th layer of the ResNet model, and the feature performance capability of the ResNet model to the interested region is enhanced;
step 4-2: changing the last single full-connection layer of the ResNet model into two parallel full-connection layers, wherein the two parallel full-connection layers respectively correspond to threat type identification and guidance type identification of an air combat threat target;
taking an output characteristic diagram of the ResNet model as the input of the two parallel full-connection layers, and respectively obtaining the final identification output of the threat type and the guidance type of the air war threat target after mapping the two full-connection layers;
and 5: taking the enhanced air combat threat target data set obtained in the step 3 as an input of an improved ResNet model, taking the label obtained in the step 2 as an air combat threat target data label, and training the improved ResNet model to obtain a trained improved ResNet model;
step 6: and inputting the air combat threat target data into the trained improved ResNet model to obtain an air combat threat target type identification result.
Further, the attention module comprises two parallel modules, a channel attention module and a space attention module;
in the channel attention module, a feature map firstly passes through a maximum pooling layer, then respectively passes through a hidden layer with two fully-connected layers, and is activated by a Relu function, then passes through a hidden layer with two fully-connected layers again, and finally passes through a Sigmoid function to obtain a one-dimensional channel attention feature map;
in the spatial attention module, the feature map passes through two channels, namely a maximum pooling layer and an average pooling layer, and then passes through a full-connection layer, the vectors of the two channels are subjected to dimensionality combination to obtain a high-dimensional vector map, then the convolution layer is connected to further extract features, and finally a final spatial attention feature map is obtained through a Sigmoid function.
Further, the channel attention module and the spatial attention module are represented as:
Mc(F)＝σ(MLP(MaxPool(F))) (3)
wherein F is a feature diagram, F 'is a channel attention mechanism, F' is a space attention mechanism, sigma is a Sigmoid function,
for dot multiplication, Mc(F) For a one-dimensional channel attention map, Ms(F) For two-dimensional spatial attention, MLP is a hidden layer structured as a fully connected layer, f7*7Is a rolled layer and a rollThe size of the nuclei, AvgPool is the average pooling layer, MaxPool is the maximum pooling layer.
Further, N is 8.
The invention has the following beneficial effects:
1. the feature extraction mechanism based on the attention module designed by the method of the invention fully utilizes the channel attention and the space attention mechanism to guide the transmission of important feature information, enhances the extraction of high-weight features in the feature vector, practically and effectively improves the perception capability of the recognition model for global information, and effectively improves the recognition accuracy of the air combat threat target type.
2. The feature mapping mechanism based on the parallel full-connection layer designed by the method can accelerate the convergence of the model, and compared with the traditional serial full-connection structure which carries out feature mapping from a single angle, the parallel full-connection layer structure can carry out different feature mapping from a plurality of angles, so that more feature information is reserved, and the classification precision of the air war threat target types is improved.
3. The improved ResNet neural network recognition model provided by the invention has good generalization and universality, and can be widely applied to real-time automatic recognition of air war threat target types and other actual scenes.
Drawings
Fig. 1 is a schematic structural diagram of an air combat threat target type identification model based on a ResNet neural network.
FIG. 2 is a schematic diagram of an identification model module of the present invention, FIG. 2(a) is a schematic diagram of a residual module structure, and FIG. 2(b) is a schematic diagram of a convolution module.
Fig. 3 is a schematic view of an attention module according to the present invention, fig. 3(a) is a schematic view of a channel attention module, and fig. 3(b) is a schematic view of a spatial attention module.
Detailed Description
The invention is further illustrated with reference to the following figures and examples.
With the continuous improvement of the autonomy level of the air combat threat target, the autonomy of the air combat threat target is improved, the autonomy of the air combat threat target in target discovery, target identification and target attack is improved, the degree of dependence on information guarantee is greatly reduced, the intellectualization and the anti-interference capability are greatly enhanced, and the advanced air combat threat target (such as an air-to-air/surface-to-air missile) becomes the maximum threat of an airborne platform. The method is used as a main basis for selecting bait types by a defense decision algorithm, and the accuracy of air war threat target type identification directly influences the effectiveness of defense decisions. Aiming at the specific requirements of a defense decision at the tail end of a fighter for identifying the type of the air combat threat target, the air combat threat target category and characteristic model are researched and established, and the air combat threat target type comprehensive identification process is analyzed. The method is characterized in that a neural network-based air combat threat target type recognition algorithm is researched by combining with the characteristics of a neural network, and under the condition of complex situation, compared with the traditional air combat threat target type recognition algorithm, a more accurate recognition result with higher confidence coefficient is obtained.
To sum up, in order to improve the accuracy of identifying the air war threat target type, the problems to be solved are mainly: how to effectively improve the accuracy of target type identification. The measures taken are as follows: (1) a ResNet network is selected as a main convolutional neural network, and an attention mechanism is added into a backbone network of a Resnet model, so that the characteristic expression capability of an interested area is enhanced, the influence of useless information such as noise on an identification result is reduced, the area focusing capability of a detection model is improved, and the identification precision of a total target is effectively improved; (2) the method comprises the steps that 2 full connection layers are arranged at the tail of a ResNet model, the two full connection layers belong to a parallel relation and respectively point to a threat type and a guidance type of an air combat threat target, namely, a unified main network is adopted for feature extraction, and different full connection layers are arranged according to different target category attributes when the features are mapped into target categories, so that the identification accuracy of the target is improved.
As shown in fig. 1, a convolutional neural network-based air war threat target identification method includes the following steps:
step 1: acquiring air combat threat target data to be identified, and constructing an air combat threat target data set;
step 2: assigning two labels to each air combat threat target data in the air combat threat target data set, wherein the labels are respectively a threat type label and a guidance type label of the air combat threat target, and the labels are adopted to distinguish different types of air combat threat targets;
and step 3: carrying out data link on 8 data of the same air combat threat target in the air combat threat target data set at different moments to obtain an enhanced air combat threat target data set;
and 4, step 4: constructing an improved ResNet model;
step 4-1: attention modules are added after the 7 th layer, the 15 th layer, the 27 th layer and the 33 th layer of the ResNet model, and the feature performance capability of the ResNet model to the interested region is enhanced;
step 4-2: changing the last single full-connection layer of the ResNet model into two parallel full-connection layers, wherein the two parallel full-connection layers respectively correspond to threat type identification and guidance type identification of an air combat threat target;
taking an output characteristic diagram of the ResNet model as the input of the two parallel full-connection layers, and respectively obtaining the final identification output of the threat type and the guidance type of the air war threat target after mapping the two full-connection layers;
and 5: taking the enhanced air combat threat target data set obtained in the step 3 as an input of an improved ResNet model, taking the label obtained in the step 2 as an air combat threat target data label, and training the improved ResNet model to obtain a trained improved ResNet model;
step 6: and inputting the air combat threat target data into the trained improved ResNet model to obtain an identification result of the air combat threat target type (including the threat type and the guidance type).
In fig. 3, the convolution-based attention module is designed. Generally, the attention mechanism in deep learning is used to remove redundant information and select information that is more important to the current target. The attention mechanism can effectively learn the weight distribution of different parts on input data or a characteristic diagram, reduce the influence caused by noise or useless information, and improve the identification capability and robustness of the model. As shown in fig. 2, the residual attention network constructs a network by using a residual mechanism, and the depth of the network is ensured while the attention mechanism is introduced. The convolution attention module simultaneously utilizes the channel information and the space information of the feature map to design an attention module, so that the model can focus on more useful information, and the data classification capability of the model is further enhanced. The attention module is designed by referring to a double-channel attention mechanism of a convolution attention module, and comprises a channel attention module, a characteristic diagram passes through a maximum pooling layer and then passes through a hidden layer with two full-connection layers, the hidden layer shares characteristics, and finally a one-dimensional channel attention diagram is obtained after passing through a Sigmiod function; for the spatial attention module, the feature map respectively passes through two paths of a maximum pooling layer and an average pooling layer, and then passes through a full connection layer sharing features, different from the channel attention module, vectors of the two paths are combined in dimensionality, after a higher-dimensional vector is obtained, the convolution layer is connected for further feature extraction, and a final spatial attention map is obtained through a Sigmoid function.
The channel attention module and the spatial attention module may be expressed as:
Mc(F)＝σ(MLP(MaxPool(F))) (3)
wherein F is a feature diagram, F 'is a channel attention mechanism, F' is a space attention mechanism, sigma is a Sigmoid function,
for dot multiplication, Mc(F) For a one-dimensional channel attention map, Ms(F) For two-dimensional spatial attention, MLP is a hidden layer structured as a fully connected layer, f7*7For convolution layer and convolution kernel sizes, AvgPool is the average pooling layer and MaxPool is the maximum poolAnd (7) layering.
The final parallel full-connection layer structure of the model can reduce the dimension of the information space of the air war threat target identification task, is favorable for the model to better perform independent identification aiming at different label types, and is favorable for the model to accelerate the convergence speed. Specifically, compared with the conventional method for identifying the type of the air war threat target, in which the types of the air war threat target are summarized as tag types such as "air-air radar guidance", "air-air infrared guidance", "ground air/ship air radar guidance", and the like, it can be seen that although the tag types are different, the similarity between the tag types is large, for example: the air-air radar guidance and the air-air infrared guidance belong to the air-air type in the threat type, and the air-air radar guidance and the ground-air/ship-air radar guidance belong to the radar guidance type in the guidance type. In view of this, there is a great similarity between these two sets of data, which is not favorable for the neural network model to accurately identify them. Therefore, the multi-parallel full-connection layer structure adopted by the invention can respectively and independently identify the threat type and the guidance type of the air combat threat target in a targeted manner, reduce the redundancy phenomenon of data in an information space caused by the arrangement of data labels, and simultaneously reduce the dimension of the information space of the data and weaken the similarity among the data of different labels in a data set, thereby accelerating the convergence speed of the model and effectively improving the identification precision of the neural network model.
Fig. 1 is a general block diagram of the identification of the object type of air war threats. Firstly, acquiring air combat threat target data to be identified and splicing different sensor data of the same air combat threat target at multiple moments together; and then the acquired air combat threat target data is used as input of a pre-trained neural network model based on ResNet air combat threat target type recognition, the neural network model based on ResNet air combat threat target type recognition carries out feature extraction on the air combat threat target data, the extracted features are respectively input into two full connection layers at the tail part, and after mapping of the full connection layers, recognition results of guidance types and threat types of the air combat threat target are output. The pre-trained training sample for identifying the neural network model based on the ResNet air combat threat target type is respectively marked with two classification labels, wherein one type is a guidance type label and comprises radar guidance and infrared guidance; the second is threat type labels, including air-air, ground-air/ship-air and non-missile. The neural network model based on ResNet air combat threat target type recognition comprises a convolutional layer module, an attention mechanism module, a parallel full-link layer module and a loss function calculation module, wherein the convolutional layer module is used for extracting characteristics of input data of a neural network, the attention mechanism module is used for adaptively improving the proportion of a certain amount of characteristic values in characteristic vectors in forward calculation, the characteristic values belong to characteristic values which greatly contribute to the classification result of the air combat threat target type, and the loss function calculation module is used for determining parameters of the neural network model based on the ResNet air combat threat target type recognition based on the marks of training samples and a plurality of type recognition results, so that the neural network model based on the ResNet air combat threat target type recognition can output corresponding air combat threat target type recognition results aiming at air combat threat target detection data to be recognized. The attention mechanism feature extraction module is utilized, channel attention and space attention are fully combined to guide the transfer of feature information, the extraction of high-weight features in feature vectors is enhanced, the expression capability of the feature information is improved, the grabbing capability of a detection model for key information is effectively improved, and the identification accuracy of the air combat threat target type is effectively improved; the method for mapping and identifying the output characteristic diagram of the main neural network by using the multi-parallel full-connection layer structure can reduce the dimension of the information space of data, and can specifically and accurately identify the threat type and the guidance type of the air combat threat target respectively.
The specific embodiment is as follows:
in order to verify the effectiveness of the algorithm, 5000 pieces of air combat threat target data at different moments are generated through a simulation system, and the algorithm is verified.
In order to verify the effectiveness of the algorithm before and after improvement, the original ResNet neural network model and the improved ResNe neural network model in the invention are respectively used for verification on the air combat threat target data set, and the experimental results are shown in Table 1. The table shows that under the same data set, the accuracy of the improved ResNet neural network model is improved to a certain extent compared with the accuracy of the original ResNet neural network model, and the improved ResNet neural network model has better feature extraction capability, so that the accuracy of the deep learning detection network in identifying the air combat threat target is effectively improved. Therefore, the improved ResNet neural network model provided by the invention effectively improves the identification rate of the air war threat target type, and the method is very effective.
TABLE 1 comparison of ResNet and modified ResNet on air combat threat target data sets
Claims (4)
1. An air combat threat target identification method based on a convolutional neural network is characterized by comprising the following steps:
step 1: acquiring air combat threat target data to be identified, and constructing an air combat threat target data set;
step 2: assigning two labels to each air combat threat target data in the air combat threat target data set, wherein the labels are respectively a threat type label and a guidance type label of the air combat threat target, and the labels are adopted to distinguish different types of air combat threat targets;
and step 3: carrying out data link on N data of the same air combat threat target in the air combat threat target data set at different times to obtain an enhanced air combat threat target data set;
and 4, step 4: constructing an improved ResNet model;
step 4-1: attention modules are added after the 7 th layer, the 15 th layer, the 27 th layer and the 33 th layer of the ResNet model, and the feature performance capability of the ResNet model to the interested region is enhanced;
step 4-2: changing the last single full-connection layer of the ResNet model into two parallel full-connection layers, wherein the two parallel full-connection layers respectively correspond to threat type identification and guidance type identification of an air combat threat target;
taking an output characteristic diagram of the ResNet model as the input of the two parallel full-connection layers, and respectively obtaining the final identification output of the threat type and the guidance type of the air war threat target after mapping the two full-connection layers;
and 5: taking the enhanced air combat threat target data set obtained in the step 3 as an input of an improved ResNet model, taking the label obtained in the step 2 as an air combat threat target data label, and training the improved ResNet model to obtain a trained improved ResNet model;
step 6: and inputting the air combat threat target data into the trained improved ResNet model to obtain an air combat threat target type identification result.
2. The convolutional neural network-based air war threat target identification method of claim 1, wherein the attention module comprises two parallel modules, a channel attention module and a spatial attention module;
in the channel attention module, a feature map firstly passes through a maximum pooling layer, then respectively passes through a hidden layer with two fully-connected layers, and is activated by a Relu function, then passes through a hidden layer with two fully-connected layers again, and finally passes through a Sigmoid function to obtain a one-dimensional channel attention feature map;
in the spatial attention module, the feature map passes through two channels, namely a maximum pooling layer and an average pooling layer, and then passes through a full-connection layer, the vectors of the two channels are subjected to dimensionality combination to obtain a high-dimensional vector map, then the convolution layer is connected to further extract features, and finally a final spatial attention feature map is obtained through a Sigmoid function.
3. The convolutional neural network-based air war threat target identification method of claim 2, wherein the channel attention module and the spatial attention module are represented as:
Mc(F)＝σ(MLP(MaxPool(F))) (3)
wherein F is a feature diagram, F 'is a channel attention mechanism, F' is a space attention mechanism, sigma is a Sigmoid function,
for dot multiplication, Mc(F) For a one-dimensional channel attention map, Ms(F) For two-dimensional spatial attention, MLP is a hidden layer structured as a fully connected layer, f7*7For convolution layer and convolution kernel sizes, AvgPool is the average pooling layer and MaxPool is the maximum pooling layer.
4. The convolutional neural network-based air war threat target identification method of claim 1, wherein N-8.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111134920.9A CN113902974B (en) | 2021-09-27 | Air combat threat target identification method based on convolutional neural network |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111134920.9A CN113902974B (en) | 2021-09-27 | Air combat threat target identification method based on convolutional neural network |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN113902974A true CN113902974A (en) | 2022-01-07 |
| CN113902974B CN113902974B (en) | 2024-05-17 |
Family
ID=
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117073848A (en) * | 2023-10-13 | 2023-11-17 | 中国移动紫金(江苏)创新研究院有限公司 | Temperature measurement method, device, equipment and storage medium |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20070260570A1 (en) * | 2006-03-28 | 2007-11-08 | Nasa Hq's. | Systems, methods and apparatus for autonomic safety devices |
| CN110472296A (en) * | 2019-07-22 | 2019-11-19 | 中北大学 | A kind of air objective intimidation estimating method based on the full connection residual error network of standardization |
| WO2021073528A1 (en) * | 2019-10-18 | 2021-04-22 | 华中光电技术研究所（中国船舶重工集团有限公司第七一七研究所） | Intelligent decision-making method and system for unmanned surface vehicle |
| US20210273953A1 (en) * | 2018-02-20 | 2021-09-02 | Darktrace Holdings Limited | ENDPOINT AGENT CLIENT SENSORS (cSENSORS) AND ASSOCIATED INFRASTRUCTURES FOR EXTENDING NETWORK VISIBILITY IN AN ARTIFICIAL INTELLIGENCE (AI) THREAT DEFENSE ENVIRONMENT |
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20070260570A1 (en) * | 2006-03-28 | 2007-11-08 | Nasa Hq's. | Systems, methods and apparatus for autonomic safety devices |
| US20210273953A1 (en) * | 2018-02-20 | 2021-09-02 | Darktrace Holdings Limited | ENDPOINT AGENT CLIENT SENSORS (cSENSORS) AND ASSOCIATED INFRASTRUCTURES FOR EXTENDING NETWORK VISIBILITY IN AN ARTIFICIAL INTELLIGENCE (AI) THREAT DEFENSE ENVIRONMENT |
| CN110472296A (en) * | 2019-07-22 | 2019-11-19 | 中北大学 | A kind of air objective intimidation estimating method based on the full connection residual error network of standardization |
| WO2021073528A1 (en) * | 2019-10-18 | 2021-04-22 | 华中光电技术研究所（中国船舶重工集团有限公司第七一七研究所） | Intelligent decision-making method and system for unmanned surface vehicle |
Non-Patent Citations (3)
| Title |
|---|
| 俞卫树;姜长生;: "基于Elman神经网络的空战威胁排序研究", 电光与控制, no. 08, 15 August 2008 (2008-08-15) * |
| 张智睿;李锵;关欣;: "密集挤压激励网络的多标签胸部X光片疾病分类", 中国图象图形学报, no. 10, 16 October 2020 (2020-10-16) * |
| 翟翔宇;杨风暴;吉琳娜;李书强;吕红亮;: "标准化全连接残差网络空战目标威胁评估", 火力与指挥控制, no. 06, 15 June 2020 (2020-06-15) * |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117073848A (en) * | 2023-10-13 | 2023-11-17 | 中国移动紫金(江苏)创新研究院有限公司 | Temperature measurement method, device, equipment and storage medium |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110516561B (en) | SAR image target identification method based on DCGAN and CNN | |
| CN109919108B (en) | Remote sensing image rapid target detection method based on deep hash auxiliary network | |
| CN110941794B (en) | Challenge attack defense method based on general inverse disturbance defense matrix | |
| CN110109059B (en) | Radar radiation source signal identification method based on deep learning network | |
| Zhu et al. | Toward understanding and boosting adversarial transferability from a distribution perspective | |
| CN111126134A (en) | Radar radiation source deep learning identification method based on non-fingerprint signal eliminator | |
| CN110569738A (en) | natural scene text detection method, equipment and medium based on dense connection network | |
| CN109655815A (en) | Sonar target detection method based on SSD | |
| CN112446357B (en) | SAR automatic target recognition method based on capsule network | |
| CN112215296B (en) | Infrared image recognition method based on transfer learning and storage medium | |
| CN113806746A (en) | Malicious code detection method based on improved CNN network | |
| CN114255403A (en) | Optical remote sensing image data processing method and system based on deep learning | |
| CN112926661A (en) | Method for enhancing image classification robustness | |
| Wang et al. | An improved crop disease identification method based on lightweight convolutional neural network | |
| CN112598032B (en) | Multi-task defense model construction method for anti-attack of infrared image | |
| CN107273914B (en) | Efficient fusion identification method based on information source self-adaptive dynamic selection | |
| CN110969203B (en) | HRRP data redundancy removing method based on self-correlation and CAM network | |
| CN113902974B (en) | Air combat threat target identification method based on convolutional neural network | |
| CN113902974A (en) | Air combat threat target identification method based on convolutional neural network | |
| Wang et al. | Learning embedding features based on multisense-scaled attention architecture to improve the predictive performance of air combat intention recognition | |
| CN116778225A (en) | SAR true and false target identification and target recognition method based on decoupling and reconstruction learning | |
| CN115496987A (en) | Small sample SAR target identification method based on deep double-similarity measurement network | |
| CN113449865B (en) | Optimization method for enhancing training artificial intelligence model | |
| CN115661576A (en) | Method for identifying airplane group intention under sample imbalance | |
| CN113409351B (en) | Unsupervised field self-adaptive remote sensing image segmentation method based on optimal transmission |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant |