WO2024082374A1 - Few-shot radar target recognition method based on hierarchical meta transfer - Google Patents
Few-shot radar target recognition method based on hierarchical meta transfer Download PDFInfo
- Publication number
- WO2024082374A1 WO2024082374A1 PCT/CN2022/133980 CN2022133980W WO2024082374A1 WO 2024082374 A1 WO2024082374 A1 WO 2024082374A1 CN 2022133980 W CN2022133980 W CN 2022133980W WO 2024082374 A1 WO2024082374 A1 WO 2024082374A1
- Authority
- WO
- WIPO (PCT)
- Prior art keywords
- samples
- category
- meta
- sample
- feature
- Prior art date
Links
- 238000000034 method Methods 0.000 title claims abstract description 41
- 238000012546 transfer Methods 0.000 title claims abstract description 21
- 238000012549 training Methods 0.000 claims abstract description 37
- 230000007246 mechanism Effects 0.000 claims abstract description 22
- 230000006870 function Effects 0.000 claims description 31
- 238000013461 design Methods 0.000 claims description 14
- 239000013598 vector Substances 0.000 claims description 13
- 230000009466 transformation Effects 0.000 claims description 12
- 238000013528 artificial neural network Methods 0.000 claims description 9
- 239000000284 extract Substances 0.000 claims description 7
- 238000013507 mapping Methods 0.000 claims description 6
- 238000004364 calculation method Methods 0.000 claims description 5
- 238000013508 migration Methods 0.000 claims description 5
- 230000005012 migration Effects 0.000 claims description 5
- 230000003068 static effect Effects 0.000 claims description 3
- 238000009826 distribution Methods 0.000 abstract description 2
- 238000002474 experimental method Methods 0.000 description 5
- 238000012360 testing method Methods 0.000 description 5
- 238000011176 pooling Methods 0.000 description 4
- 238000004422 calculation algorithm Methods 0.000 description 3
- 238000000605 extraction Methods 0.000 description 3
- 230000004913 activation Effects 0.000 description 2
- 238000009827 uniform distribution Methods 0.000 description 2
- 102100024506 Bone morphogenetic protein 2 Human genes 0.000 description 1
- 101000762366 Homo sapiens Bone morphogenetic protein 2 Proteins 0.000 description 1
- 238000013473 artificial intelligence Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000013527 convolutional neural network Methods 0.000 description 1
- 230000007423 decrease Effects 0.000 description 1
- 238000013135 deep learning Methods 0.000 description 1
- 238000013136 deep learning model Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000008034 disappearance Effects 0.000 description 1
- 230000010365 information processing Effects 0.000 description 1
- 239000000463 material Substances 0.000 description 1
- 230000010287 polarization Effects 0.000 description 1
- 230000008569 process Effects 0.000 description 1
- 238000012545 processing Methods 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 238000004088 simulation Methods 0.000 description 1
- 230000001360 synchronised effect Effects 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/10—Terrestrial scenes
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N20/00—Machine learning
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/20—Image preprocessing
- G06V10/26—Segmentation of patterns in the image field; Cutting or merging of image elements to establish the pattern region, e.g. clustering-based techniques; Detection of occlusion
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/70—Arrangements for image or video recognition or understanding using pattern recognition or machine learning
- G06V10/764—Arrangements for image or video recognition or understanding using pattern recognition or machine learning using classification, e.g. of video objects
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/70—Arrangements for image or video recognition or understanding using pattern recognition or machine learning
- G06V10/77—Processing image or video features in feature spaces; using data integration or data reduction, e.g. principal component analysis [PCA] or independent component analysis [ICA] or self-organising maps [SOM]; Blind source separation
- G06V10/774—Generating sets of training patterns; Bootstrap methods, e.g. bagging or boosting
Definitions
- the invention belongs to the technical field of radar target recognition, and in particular relates to a small sample radar target recognition method based on hierarchical element migration.
- Radar target recognition technology refers to the technology of using radar to detect targets and determine the type, model and other attributes of the target by analyzing the captured information. It shows great application potential in fields such as terrain exploration and battlefield reconnaissance.
- artificial intelligence technology deep learning methods have attracted widespread attention from researchers due to their automatic and powerful feature extraction capabilities, which has promoted the emergence and advancement of intelligent radar target recognition technology.
- deep learning model training often relies on a large number of labeled samples. Due to timeliness constraints and resource limitations, obtaining a large number of labeled samples consumes huge manpower, material resources and time costs. Therefore, using meta-learning to share knowledge in small sample scenarios to improve target recognition performance is one of the current research hotspots in the field of radar target recognition technology.
- the purpose of the present invention is to provide a small sample radar target recognition method based on hierarchical meta-transfer to overcome the above-mentioned shortcomings.
- the present invention extracts features based on the attention mechanism, and hierarchical deep knowledge transfer at the feature level, sample level, and task level to seek an embedding space that makes the sample close to the category atoms of the same type of target and far away from the category atoms of other types of targets.
- a feature encoder based on the attention mechanism is designed at the feature level to fully exploit the global domain-invariant features of the sample to overcome the domain difference problem of the sample in the data distribution; an atom encoder is designed at the sample level to generate more stable category atoms to avoid the influence of outlier samples; at the task level, a meta-learner is designed to accumulate the learning experience of the training task and transfer it to the new task, cultivate the model's ability to transfer knowledge across tasks, and realize meta-transfer target recognition. Therefore, the small sample radar target recognition method based on hierarchical meta-transfer proposed in the present invention is an intelligent target recognition method.
- a small sample radar target recognition method based on hierarchical element migration includes the following steps:
- P is the total number of tasks, It includes support set and query set, where the support set is composed of labeled samples extracted from the source domain, and the query set is composed of labeled samples extracted from the target domain;
- the deep global features of the query set samples and the distances between atoms of different categories are used to obtain the probability that the corresponding samples belong to different categories.
- the meta-learner loss function is designed based on the probability, and the meta-learner is updated by minimizing the loss function to obtain the updated meta-learner.
- step S4 Complete all training tasks by repeating step S3 to obtain the meta-learner trained by all meta-training tasks.
- the trained meta-learner is recorded as
- the labeled samples of the task to be tested are the support set, and the unlabeled samples to be tested are the query set; the meta-learner obtained in S4 is used for initialization
- a feature encoder for target recognition and a category atom encoder are obtained, and the feature encoder for target recognition is used to extract deep global features for the support set and query set samples.
- the category atom encoder for target recognition is used to calculate and update the category atoms based on the deep global features of the support set, and the distance function dist( ⁇ ) is used to calculate the distance between the deep global features of the sample to be tested in the query set and the atoms of different categories, and the label of the category atom with the closest distance is selected as the predicted label of the sample to be tested to obtain the recognition result.
- the support set is constructed by extracting labeled samples in the source domain in the form of K way N shot, which is defined as K way N shot means randomly extracting N labeled training samples from each category of K types of targets. is the nth sample of the kth class target; the query set is composed of labeled samples extracted in the target domain in the form of K way M shot, defined as in, is the mth sample of the kth class target; the samples in the support set and query set are samples of the same class target in different domains, and the corresponding class labels are defined as in,
- the feature encoder It includes a neural network module and an attention mechanism module.
- the specific method of extracting deep global features is as follows:
- the generalized features are divided into blocks and straightened into vectors.
- the dimension of each vector is d 1 , denoted as [b 1 , b 2 , ..., b R ] T , where R is the number of blocks.
- a learnable vector b 0 of the same dimension is added to represent the global features of the entire sample.
- Feature B is first mapped to a high-dimensional space through a fully connected layer, and the dimension of the high-dimensional space is recorded as d2 , and then mapped back to a low-dimensional space of d1 to obtain the deep feature
- d2 the dimension of the high-dimensional space
- d1 the dimension of the high-dimensional space
- the support set and query set are feature encoded to obtain: in,
- the tasks are Deep global features of the support and query sets, and
- step S32 the specific method of updating the category atom encoder and the category atom is:
- sample-level global features are transformed back to d1 dimensions through linear mapping LN( ⁇ ), and the residual structure and deep global features are combined to obtain
- the features are first Map to a high-dimensional space of d2 , and then map back to a low-dimensional space of d1 to obtain deep features With features
- the residual structure is used to combine and obtain sample-level deep global features
- sample-level deep global features are averaged to obtain the sample-level category atoms
- step S33 the specific method for updating the meta-learner is:
- margin is the set threshold
- ⁇ is the balance parameter
- the meta-learner is updated by minimizing the loss function to obtain the updated meta-learner
- the beneficial effects of the present invention are as follows: for small sample target recognition scenarios, the present invention fully mines the global features of samples at the feature level, fully explores the robustness features of different samples of the same target at the sample level, and designs a meta-learner at the task level to effectively accumulate learning experience of different tasks.
- the quality of feature information is improved, the negative impact of outlier samples is reduced, the autonomous learning ability of the model is cultivated, and the robustness of small sample target recognition technology is improved.
- the small sample radar target recognition method based on hierarchical meta-transfer proposed by the present invention is an intelligent radar target recognition method.
- FIG. 1 is a flow chart of an algorithm of the present invention.
- FIG. 2 is a comparison chart of the recognition accuracy of the background technology method and the method of the present invention.
- the present invention designs a small sample radar target recognition method based on hierarchical meta-transfer, including feature level, sample level and task level.
- an attention mechanism is used to construct a feature encoder to extract more important features in a single sample;
- an attention mechanism is used to construct an atom encoder, and high-quality category atoms are generated as representative information of the corresponding category by integrating the information of different samples of the same type of target.
- a meta-learner is constructed to acquire autonomous learning ability by accumulating learning experience of different meta-training tasks.
- the trained meta-learner is further optimized based on a small number of labeled samples to generate high-quality category atoms for target recognition.
- the sample to be tested is compared with the category atom, and the category of the category atom with the highest similarity is selected as the predicted category of the test sample to complete the recognition of the test sample.
- This example is a practical application of the method according to the present invention.
- synchronous initialization is performed when establishing the feature encoder and the category atom encoder so that they can be processed faster.
- Step 1 Collect and preprocess original image samples in the source domain and target domain respectively, and preliminarily filter out redundant information of the target background to prepare for training the model.
- the radar obtains the original images of each target at different pitch angles when it is static. At each fixed pitch angle, the target is observed at different azimuth angles. The acquired images are recorded as source domain and target domain according to the different pitch angles, and they are cut and preprocessed.
- Step 2 Use samples to build training tasks
- Each task includes a support set and a query set to train an object recognition model with autonomous learning capabilities.
- K k classification task
- the support set is composed of labeled samples extracted from the source domain in the form of K way N shot and recorded as Among them, K way N shot means randomly extracting N labeled training samples from each category of K types of targets. is the nth sample of the kth class target; the labeled samples are extracted in the target domain in the form of K way M shot to form a query set and recorded as in, is the mth sample of the kth class target.
- the samples in the support set and query set should be samples of the same class target in different domains.
- the corresponding class labels are in,
- Step 3 In order to accumulate learning experience from different tasks and cultivate the model's ability to learn autonomously, the meta-learner is trained and learned through the hierarchical meta-transfer model.
- the hierarchical meta-transfer model is composed of feature level, sample level and task level, specifically:
- Step 31 Design feature encoder at feature level For the training task obtained in step 2
- the support set and query set are respectively extracted with features to explore the deep information of the sample for identification. Further, the specific steps of step 31 are:
- Step 31-1 Design feature encoder at feature level
- the feature encoder consists of a neural network module and an attention mechanism module.
- the neural network module has a strong feature extraction capability and can mine the deep features of the sample.
- the attention mechanism module is to enable the model to selectively focus on the important information in the sample and improve the efficiency of the model's information processing.
- Step 31-2 Use the neural network module and attention mechanism to extract the deep global features of the sample.
- the specific steps are as follows:
- Step 31-2-1 Use the convolutional neural network module conv( ⁇ ) to train the support set samples Extract generalized features.
- the support set sample representation symbol is abbreviated as S.
- the feature extraction process is as follows:
- Step 31-2-2 Divide the sample generalization features obtained in step 31-2-1 into blocks and straighten them into vectors.
- the dimension of each vector is d 1 .
- All vectors are recorded as [b 1 , b 2 , ..., b R ] T , where R is the number of blocks.
- Step 31-2-3 To further filter out redundant information, the feature B obtained in step 31-2-2 is transformed and reduced to different d-dimensional embedding subspaces:
- V BW v (4)
- Step 31-2-4 To alleviate the gradient disappearance, the global features obtained in step 31-2-3 are transformed back to d 1 dimensions through linear mapping LN( ⁇ ), and the residual structure is used to combine with the features obtained in step 31-2-2:
- Step 31-3 Since the information in high-dimensional space is richer, a layer of fully connected network is used to map the features obtained in step 31-2 to the high-dimensional space. Note that the dimension of the high-dimensional space is d2 , and then a layer of fully connected network is used to map it back to the original dimension d1 .
- Each fully connected layer is processed with an activation function to learn and obtain more abstract deep features. Enhance the expressiveness of information. To avoid the gradient vanishing problem, combine it with the features obtained in step 3-2 using a residual structure to obtain a deep global feature:
- Step 31-4 Task The support set and query set are feature encoded: in, The tasks are Deep global features of the support and query sets, and
- Step 32 Design an attention-based category atom encoder at the sample level And in the current training task The updated category atoms are calculated to provide reliable representative information for target recognition. Further, the specific steps of step 4 are as follows:
- Step 32-1 For the task Designing category atom encoders at the sample level And using the current meta-learner Class Atom Encoder in Initialize it:
- Step 32-2 Use the category atom encoder obtained in step 32-1
- the deep global feature calculation task for the support set samples obtained in step 31 The specific steps are as follows:
- Step 32-2-1 To remove redundant information, explore the deep features of samples in different embedding subspaces and analyze the deep global features of the support set samples. Transform and reduce the dimension to d dimension respectively:
- Step 32-2-2 To alleviate the gradient vanishing, the sample-level global features obtained in step 32-2-1 are transformed back to d 1 dimensions through linear mapping LN( ⁇ ), and the residual structure is combined with the support set deep global features obtained in step 31:
- Step 32-2-3 Since the information in high-dimensional space is richer, a layer of fully connected network is used to map the features obtained in step 32-2-2 to a high-dimensional space of d2 dimensions, and then a layer of fully connected network is used to map it back to the original dimension of d1 dimensions.
- Each layer of fully connected layer is processed with an activation function to learn and obtain more abstract deep features. Enhance the expressiveness of information. To avoid the gradient vanishing problem, combine it with the features obtained in step 32-2-2 using a residual structure to obtain sample-level deep global features:
- Step 32-2-4 Average the sample-level deep global features obtained in step 32-2-3 to obtain the category atoms after the sample-level attention mechanism exploration
- Step 32-2-5 Calculation task All the class atoms in are represented as in Corresponding to the processing flow of step 32-2-1 to step 32-2-4.
- Step 32-3 Calculate the task obtained in step 31
- the deep global features of the support set samples and the distances of different types of atoms obtained in step 32-2 further obtain the samples
- the probability of being judged as category k is:
- dist( ⁇ ) is the distance function
- Step 32-4 Design and minimize the category atom loss function according to probability to update the category atom encoder and category atoms. The specific steps are as follows:
- Step 32-4-1 Design the following loss function so that the sample The probability of being judged as category k is as large as possible to obtain a model with recognition ability. Minimize the loss function and update the category atom encoder:
- Step 32-4-2 Note that the updated model is The updated category atom is in,
- Step 33 Accumulate the learning experience of the current training task at the task level and update the meta-learner to In order to enable the meta-learner to have autonomous learning capabilities to cope with new target recognition tasks, further, the specific steps of step 33 are:
- Step 33-1 Calculate the task obtained in step 31
- the deep global features of the query set samples and the distances of atoms of different categories obtained in step 32 are further obtained.
- the probability of being judged as category k is:
- dist( ⁇ ) is the distance function
- Step 33-2 Design a meta-learner loss function based on probability, minimize the loss function to update the meta-learner, and obtain The specific steps are:
- Step 33-2-1 Design the meta-learner classification loss function based on the classification probability obtained in step 33-1:
- Step 33-2-2 In order to improve the separability of samples and enhance the recognition performance of the model, the model training also uses contrast loss as the loss function, which is defined as follows:
- margin is the set threshold. This constraint can reduce the sample characteristics With the corresponding class atom The distance between them increases with the distance between atoms of other types. The distance between them should be as large as possible.
- Step 33-2-3 Combine the loss functions of step 33-2-1 and step 33-2-2 to obtain the total meta-learner loss function:
- ⁇ is a balance parameter. Minimize the meta-learner loss function to update the meta-learner and obtain The updated meta-learner Thus accumulating in the task learning experience.
- Step 5 The labeled samples of the task to be tested are called the support set, and the unlabeled samples to be tested are called the query set. To identify the sample to be tested, further, the specific steps of step 5 are:
- Step 5-1 Process the task to be tested based on the learning experience accumulated on the training task, and initialize the task model to be tested according to step 31 And extract deep global features for support set and query set samples.
- Step 5-2 Initialize the task model to be tested according to step 32 Calculate and update the category atoms using the support set;
- Step 5-3 Use the distance function dist( ⁇ ) to calculate the deep global features of the query set sample and the distances between atoms of different categories, select the label of the category atom with the closest distance as the predicted label of the sample to be tested, and obtain the recognition result.
- the implementation model is used to experiment with the MSTAR dataset for acquisition and recognition of moving and stationary targets.
- the sensor of this dataset uses a high-resolution spotlight synthetic aperture radar, adopts HH polarization mode, works in the X-band, and has a resolution of 0.3m ⁇ 0.3m.
- Most of the data are SAR slice images of stationary vehicles, which contain a total of ten types of targets, namely BMP2, T72, BTR70, 2S1, BRDM2, BTR60, D7, T62, ZIL131, ZSU234 and T72. Seven types of targets are taken to form the meta-training task, and the remaining three types of targets are used to construct the task to be tested.
- the sample data observed at a pitch angle of 17° is used as the source domain sample, and the sample data observed at a pitch angle of 15° is used as the target domain sample.
- Table 1 The specific number of samples in the experiment is shown in Table 1.
- the sample image size is cut into 64 ⁇ 64 at the center.
- This case uses a 3-classification task, that is, each meta-training task and the task to be tested contains three types of targets.
- 3 of the 7 types of targets are randomly selected to form the meta-training task.
- samples are randomly extracted from the source domain in the form of 3way 5shot to form the support set of the task, that is, 5 samples are randomly extracted from each of the 3 types of targets in the source domain for this task;
- the query set is composed of samples randomly extracted from the target domain in the form of 3way 15shot, that is, 15 samples are randomly extracted from each of the 3 types of targets in this task.
- the samples in the support set and the query set are all labeled samples.
- samples of the target category to be tested are randomly extracted to form the task to be tested, where the support set comes from the source domain and is the labeled samples observed at a pitch angle of 17°, and the query set comes from the target domain and is the samples to be tested observed at a pitch angle of 15°.
- this case also simulates target domain samples under different noise environments.
- a certain percentage of pixels are randomly selected from the test samples of the query set in the test task, and the pixels are destroyed by replacing the intensity of their pixels with independent samples that obey the uniform distribution.
- the added random noise obeys the uniform distribution of [0, ⁇ max ], where ⁇ max is the maximum value of the pixels in the image.
- the selected pixel ratios are 0%, 5%, and 15%, respectively, representing the target domains under different noise environments, where 0% represents the test samples constructed from the 15° pitch angle observation samples in the original data set.
- the present invention designs experiments in different noise environments for small sample target recognition to verify the superiority of the proposed algorithm, and compares the recognition results of the background technology method and the method of the present invention on the task to be tested.
- the neural network module of the feature encoder consists of four convolutional layers, and the maximum pooling operation is used after each convolutional layer to reduce the size of the model and improve the calculation speed.
- Table 2 shows the detailed parameters of each convolutional layer and pooling operation, including the size of the convolution kernel, the step size during convolution, the padding size, and the size of the pooling window.
- the background technology methods all show a significant decline to varying degrees.
- the recognition accuracy of background technology method 1 in 0% and 15% noise environments is 77.43% and 71.66% respectively, and the recognition accuracy of the background technology method is 71.67% and 68.1%, while the method of the present invention can still maintain a high recognition rate, with recognition accuracy rates of 83.86%, 82.24%, and 81.92% in 0%, 5%, and 15% noise environments, respectively, which has obvious advantages.
- the experimental results prove that the present invention effectively explores the deep global features of samples in small sample target recognition scenarios, cultivates the autonomous learning ability of the model, establishes a more stable meta-learning model, and improves target recognition performance.
- Convolutional Layer Convolution kernel size Step Length Filling size Pooling window size level one 5 ⁇ 5 1 0 2 ⁇ 2 Second floor 3 ⁇ 3 1 0 2 ⁇ 2 the third floor 3 ⁇ 3 1 1 2 ⁇ 2 Fourth floor 3 ⁇ 3 1 1 2 ⁇ 2
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Multimedia (AREA)
- Software Systems (AREA)
- Computing Systems (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Artificial Intelligence (AREA)
- Evolutionary Computation (AREA)
- Medical Informatics (AREA)
- Databases & Information Systems (AREA)
- General Health & Medical Sciences (AREA)
- Health & Medical Sciences (AREA)
- Data Mining & Analysis (AREA)
- General Engineering & Computer Science (AREA)
- Mathematical Physics (AREA)
- Image Analysis (AREA)
Abstract
The present invention belongs to the technical field of target recognition, and particularly relates to a few-shot radar target recognition method based on hierarchical meta transfer. In the present invention, features are extracted on the basis of an attention mechanism and hierarchical deep knowledge transfer at a feature level, a sample level and a task level is performed in order to search for an embedding space, such that a sample is close to a category atom of targets of the same category and is away from a category atom of targets of other categories. At the feature level, a feature encoder based on the attention mechanism is designed and global domain-invariant representations of samples are fully mined, such that the problem of domain difference in data distribution of the samples is solved; at the sample level, an atom encoder is designed, and more stable category atoms are generated, such that the influence of outlier samples is avoided; and at the task level, a meta learner is designed, learning experience of training tasks is accumulated and is transferred to a new task, and the capability of cross-task knowledge transfer of a model is developed, such that target recognition based on meta transfer is realized. The target recognition method of the present invention is an intelligent target recognition method.
Description
本发明属于雷达目标识别技术领域,具体的说是涉及一种基于层级化元迁移的小样本雷达目标识别方法。The invention belongs to the technical field of radar target recognition, and in particular relates to a small sample radar target recognition method based on hierarchical element migration.
雷达目标识别技术是指利用雷达对目标进行探测,通过分析所捕获信息确定目标的种类、型号等属性的技术,在地形勘探及战场侦察等领域展现出很好的应用潜力。随着人工智能技术的发展,深度学习方法因其自动且强大的特征提取能力而受到研究者的广泛关注,推动了智能化雷达目标识别技术的产生与进步。然而,深度学习的模型训练往往依赖于大量的标记样本。受时效性约束和资源限制,获取大量的标记样本耗费巨大的人力、物力及时间成本。因此,在小样本场景下利用元学习进行知识共享,从而提升目标识别性能,是当前雷达目标识别技术领域的研究热点之一。Radar target recognition technology refers to the technology of using radar to detect targets and determine the type, model and other attributes of the target by analyzing the captured information. It shows great application potential in fields such as terrain exploration and battlefield reconnaissance. With the development of artificial intelligence technology, deep learning methods have attracted widespread attention from researchers due to their automatic and powerful feature extraction capabilities, which has promoted the emergence and advancement of intelligent radar target recognition technology. However, deep learning model training often relies on a large number of labeled samples. Due to timeliness constraints and resource limitations, obtaining a large number of labeled samples consumes huge manpower, material resources and time costs. Therefore, using meta-learning to share knowledge in small sample scenarios to improve target recognition performance is one of the current research hotspots in the field of radar target recognition technology.
文献“Guo J,Wang L,Zhu D,et al.SAR Target Recognition With Limited Samples Based on Meta Knowledge Transferring Using Relation Network[C]//2020International Symposium on Antennas and Propagation(ISAP).IEEE,2021:377-378”提出一种基于对比学习的小样本雷达目标识别方法,通过构建神经网络来计算两个输入样本之间的距离来分析匹配程度,从而判断其是否属于同一类。在对未标记样本进行分类时,以距离最近的标记样本的标签作为预测标签。但是,这种方法需要将待测样本与每一个标记样本进行对比,计算繁琐且复杂度较高。为解决上述问题,文献“Cai J,Zhang Y,Guo J,et al.ST-PN:A Spatial Transformed Prototypical Network for Few-Shot SAR Image Classification[J].Remote Sensing,2022,14(9):2019”提出一种基于类别原子的小样本雷达目标识别方法,将每类标记样本的特征取平均作为类别原子,在对未标记样本分类时,只需将其特征和类别原子进行对比,从而降低计算复杂度,但该方法没有对样本特征和目标特征进行深度探索,生成的类别原子在小样本场景下很容易受离群样本的影响,存在质量差,稳健性差的问题,进而影响目标识别性能。同时,考虑到这些元学习方法只是寻求样本的相似度关系,在面对和训练任务不同的新任务时,模型无法优化实现知识的跨任务迁移。因此,研究基于层级化元迁移的小样本目标识别方法有望进一步提升目标识别性能。The paper "Guo J, Wang L, Zhu D, et al. SAR Target Recognition With Limited Samples Based on Meta Knowledge Transferring Using Relation Network [C] // 2020 International Symposium on Antennas and Propagation (ISAP). IEEE, 2021: 377-378" proposes a small sample radar target recognition method based on contrastive learning. It constructs a neural network to calculate the distance between two input samples to analyze the degree of matching, so as to determine whether they belong to the same category. When classifying unlabeled samples, the label of the nearest labeled sample is used as the predicted label. However, this method requires comparing the sample to be tested with each labeled sample, which is cumbersome and complex to calculate. To solve the above problems, the paper "Cai J, Zhang Y, Guo J, et al. ST-PN: A Spatial Transformed Prototypical Network for Few-Shot SAR Image Classification [J]. Remote Sensing, 2022, 14(9): 2019" proposed a small sample radar target recognition method based on category atoms. The features of each type of labeled samples are averaged as category atoms. When classifying unlabeled samples, only their features need to be compared with category atoms, thereby reducing the computational complexity. However, this method does not conduct in-depth exploration of sample features and target features. The generated category atoms are easily affected by outlier samples in small sample scenarios, and have problems of poor quality and poor robustness, which in turn affects the target recognition performance. At the same time, considering that these meta-learning methods only seek the similarity relationship between samples, when faced with new tasks that are different from the training tasks, the model cannot optimize the cross-task transfer of knowledge. Therefore, studying small sample target recognition methods based on hierarchical meta-transfer is expected to further improve target recognition performance.
发明内容Summary of the invention
本发明的目的是,为克服上述不足,提供一种基于层级化元迁移的小样本雷达目标识别方法。本发明基于注意力机制提取特征,在特征级、样本级、和任务级上的层级化深度知识迁移,以寻求一个嵌入空间使得样本接近同类目标的类别原子,远离于其他类目标的类别原子。其中,在特征级设计了基于注意力机制的特征编码器,充分挖掘样本全局性的域不变特征,以克服样本在数据分布上的域差异问题;在样本级设计原子编码器,生成更加稳定的类别原子,以避免离群样本的影响;在任务级,设计元学习器累积训练任务的学习经验迁移至新任务,培养模型跨任务知识迁移的能力,实现元迁移目标识别。因此本发明提出的基于层级化元迁移的小样本雷达目标识别方法是一种智能的目标识别方法。The purpose of the present invention is to provide a small sample radar target recognition method based on hierarchical meta-transfer to overcome the above-mentioned shortcomings. The present invention extracts features based on the attention mechanism, and hierarchical deep knowledge transfer at the feature level, sample level, and task level to seek an embedding space that makes the sample close to the category atoms of the same type of target and far away from the category atoms of other types of targets. Among them, a feature encoder based on the attention mechanism is designed at the feature level to fully exploit the global domain-invariant features of the sample to overcome the domain difference problem of the sample in the data distribution; an atom encoder is designed at the sample level to generate more stable category atoms to avoid the influence of outlier samples; at the task level, a meta-learner is designed to accumulate the learning experience of the training task and transfer it to the new task, cultivate the model's ability to transfer knowledge across tasks, and realize meta-transfer target recognition. Therefore, the small sample radar target recognition method based on hierarchical meta-transfer proposed in the present invention is an intelligent target recognition method.
本发明的技术方案是:The technical solution of the present invention is:
一种基于层级化元迁移的小样本雷达目标识别方法,包括以下步骤:A small sample radar target recognition method based on hierarchical element migration includes the following steps:
S1、通过雷达获取各目标静态时在源域和目标域的原始图像,将对目标在不同方位角下进行观测得到的图像进行切割处理后得到样本;S1. Obtaining original images of each target in the source domain and the target domain when the target is static through radar, and cutting the images obtained by observing the target at different azimuth angles to obtain samples;
S2、利用样本构建训练任务
其中P是任务总数,任务
包括支撑集和查询集,其中支撑集是从源域中抽取标记样本构成,查询集是从目标域中抽取标记样本构成;
S2. Use samples to build training tasks Where P is the total number of tasks, It includes support set and query set, where the support set is composed of labeled samples extracted from the source domain, and the query set is composed of labeled samples extracted from the target domain;
S3、通过层级化元迁移模型进行训练学习,对元学习器
进行更新,具体为:
S3. Training and learning through hierarchical meta-transfer models Update, specifically:
S31、在特征级构建基于注意力机制的特征编码器
利用元学习器对特征编码器进行初始化
后,提取
中支撑集和查询集的深度全局性特征;
S31. Constructing a feature encoder based on the attention mechanism at the feature level Initialize the feature encoder using a meta-learner After that, extract Deep global features of the support set and query set;
S32、在样本级构建基于注意力机制的类别原子编码器
利用元学习器对类别原子编码器进行初始化
后,基于获得的
支撑集样本的深度全局性特征计算
的类别原子,根据支撑集样本和不同类别原子的距离获得对应样本归属于不同类的概 率,再根据概率设计并最小化类别原子损失函数,以此更新类别原子编码器和类别原子;
S32. Constructing a category atom encoder based on attention mechanism at sample level Initialize the category atom encoder using a meta-learner Afterwards, based on the obtained Deep global feature calculation of support set samples The category atoms of the support set are obtained, and the probability of the corresponding samples belonging to different categories is obtained according to the distance between the support set samples and the different category atoms. Then, the category atom loss function is designed and minimized according to the probability to update the category atom encoder and category atoms.
S33、在任务级累计当前训练任务的学习经验,更新元学习器:S33. Accumulate the learning experience of the current training task at the task level and update the meta-learner:
根据
查询集样本的深度全局性特征和不同类别原子的距离,获得对应样本归属于不同类的概率,根据概率设计元学习器损失函数,最小化该损失函数对元学习器进行更新,获得更新后的元学习器
according to The deep global features of the query set samples and the distances between atoms of different categories are used to obtain the probability that the corresponding samples belong to different categories. The meta-learner loss function is designed based on the probability, and the meta-learner is updated by minimizing the loss function to obtain the updated meta-learner.
S4、通过重复步骤S3完成所有训练任务,获得所有元训练任务训练出的元学习器,记训练出的元学习器为
S4. Complete all training tasks by repeating step S3 to obtain the meta-learner trained by all meta-training tasks. The trained meta-learner is recorded as
S5、记待测任务的标记样本为支撑集,未标记的待测样本为查询集;利用S4获得的元学习器进行初始化
得到目标识别用特征编码器和类别原子编码器,利用目标识别用特征编码对支撑集和查询集样本提取深度全局性特征;利用目标识别用类别原子编码器基于支撑集深度全局性特征计算并更新类别原子,利用距离函数dist(·)计算查询集中待测样本的深度全局性特征和不同类别原子的距离,选取距离最近的类别原子的标签作为待测样本的预测标签,得到识别结果。
S5: The labeled samples of the task to be tested are the support set, and the unlabeled samples to be tested are the query set; the meta-learner obtained in S4 is used for initialization A feature encoder for target recognition and a category atom encoder are obtained, and the feature encoder for target recognition is used to extract deep global features for the support set and query set samples. The category atom encoder for target recognition is used to calculate and update the category atoms based on the deep global features of the support set, and the distance function dist(·) is used to calculate the distance between the deep global features of the sample to be tested in the query set and the atoms of different categories, and the label of the category atom with the closest distance is selected as the predicted label of the sample to be tested to obtain the recognition result.
进一步的,步骤S2中,所述支撑集是通过K way N shot形式在源域中抽取标记样本构成,定义为
K way N shot指的是对K类目标每个类别随机抽取N个标记训练样本,
是第k类目标的第n个样本;查询集是通过K way M shot的形式在目标域抽取标记样本构成,定义为
其中,
是第k类目标的第m个样本;支撑集和查询集中的样本是相同类别目标在不同域中的样本,定义对应的类别标签为
其中,
Furthermore, in step S2, the support set is constructed by extracting labeled samples in the source domain in the form of K way N shot, which is defined as K way N shot means randomly extracting N labeled training samples from each category of K types of targets. is the nth sample of the kth class target; the query set is composed of labeled samples extracted in the target domain in the form of K way M shot, defined as in, is the mth sample of the kth class target; the samples in the support set and query set are samples of the same class target in different domains, and the corresponding class labels are defined as in,
进一步的,步骤S31中,所述特征编码器
包括一个神经网络模块和注意力机制模块,提取深度全局性特征的具体方式为:
Furthermore, in step S31, the feature encoder It includes a neural network module and an attention mechanism module. The specific method of extracting deep global features is as follows:
通过神经网络模块对样本提取泛化特征;Extract generalized features from samples through neural network modules;
将泛化特征分块并拉直成向量,每个向量的维度是d
1,记为[b
1,b
2,…,b
R]
T其中,R是分块的个数,添加一个同维度的可学习向量b
0代表整个样本的全局特征,记嵌入可学习信息后的特征为B=[b
0,b
1,b
2,…,b
R]
T,
The generalized features are divided into blocks and straightened into vectors. The dimension of each vector is d 1 , denoted as [b 1 , b 2 , …, b R ] T , where R is the number of blocks. A learnable vector b 0 of the same dimension is added to represent the global features of the entire sample. The features after embedding the learnable information are denoted as B = [b 0 , b 1 , b 2 , …, b R ] T .
将特征B分别进行变换处理降维到不同的d维嵌入子空间:Transform feature B and reduce its dimension to different d-dimensional embedding subspaces:
E=BW
e
E=BW e
U=BW
u
U= BWu
V=BW
v
V=BW v
其中,W
e,W
u,W
v是不同的变换矩阵,E,U,V是不同嵌入子空间中的变换特征,利用注意力机制处理获得全局性特征
Among them, We , Wu , Wv are different transformation matrices, E, U, V are the transformation features in different embedding subspaces, and the attention mechanism is used to obtain global features.
将全局性特征经过线性映射LN(·)变回d
1维,采用残差结构和特征B进行结合得到
The global features are transformed back to d1 dimensions through linear mapping LN(·), and the residual structure is combined with feature B to obtain
通过全连接层将特征B先映射到高维空间,记高维空间的维数为d
2维,再映射回 d
1维的低维空间,得到深层特征
与特征
采用残差结构进行结合得到深度全局性特征
将可学习向量
取出作为对应样本的深度全局性特征
Feature B is first mapped to a high-dimensional space through a fully connected layer, and the dimension of the high-dimensional space is recorded as d2 , and then mapped back to a low-dimensional space of d1 to obtain the deep feature With features Use residual structure to combine and obtain deep global features The learnable vector Take out the deep global features as the corresponding samples
对于任务
利用特征编码器
对任务
的支撑集和查询集进行特征编码得到:
其中,
分别是任务
支撑集和查询集的深度全局性特征,并且
For the task Using feature encoder To the task The support set and query set are feature encoded to obtain: in, The tasks are Deep global features of the support and query sets, and
进一步的,步骤S32中,所述更新类别原子编码器和类别原子的具体方法为:Furthermore, in step S32, the specific method of updating the category atom encoder and the category atom is:
利用S31中提取的支撑集深度全局性特征
将
分别进行变换处理降维到不同的d维嵌入子空间:
Using the deep global features of the support set extracted from S31 Will Transform and reduce the dimensions to different d-dimensional embedding subspaces:
其中,
和
是不同的变换矩阵,
和
是不同嵌入子空间中的变换特征,利用注意力机制得到样本级全局性特征
in, and are different transformation matrices, and It is the transformation feature in different embedding subspaces, and the sample-level global feature is obtained by using the attention mechanism
将样本级全局性特征经过线性映射LN(·)变回d
1维,采用残差结构和深度全局性特征进行结合得到
The sample-level global features are transformed back to d1 dimensions through linear mapping LN(·), and the residual structure and deep global features are combined to obtain
通过全连接层先将特征
映射到d
2维的高维空间,再映射回d
1维的低维空间,得到深层特征
与特征
采用残差结构进行结合得到样本级深度全局性特征
Through the fully connected layer, the features are first Map to a high-dimensional space of d2 , and then map back to a low-dimensional space of d1 to obtain deep features With features The residual structure is used to combine and obtain sample-level deep global features
对样本级深度全局特征取平均,得到样本级类别原子
The sample-level deep global features are averaged to obtain the sample-level category atoms
利用类别原子编码器基于得到的支撑集样本的深度全局性特征计算任务
中所有的类别原子并表示为
其中
Using the category atomic encoder to calculate the deep global features of the support set samples All the class atoms in are represented as in
根据获得的任务
支撑集样本的深度全局性特征和获得的不同类别原子的距离,获得样本
被判定为类别k的概率为:
According to the tasks obtained The deep global features of the support set samples and the distances of atoms of different categories obtained to obtain the samples The probability of being judged as category k is:
其中,dist(·)是距离函数;Where dist(·) is the distance function;
根据概率设计并最小化类别原子损失函数:Design and minimize the category atom loss function based on probability:
对类别原子编码器进行更新,记更新后的模型为
更新后的类别原子为
其中,
Update the category atom encoder and record the updated model as The updated category atom is in,
进一步的,步骤S33中,更新元学习器的具体方法为:Furthermore, in step S33, the specific method for updating the meta-learner is:
根据获得的任务
查询集样本的深度全局性特征和获得的不同类别原子的距离,获得样本
被判定为类别k的概率为:
According to the tasks obtained The deep global features of the query set samples and the distances of atoms of different categories are obtained to obtain the samples The probability of being judged as category k is:
根据概率设计元学习器损失函数:Design the meta-learner loss function based on probability:
其中,margin是设置的阈值,γ是平衡参数,通过最小化损失函数对元学习器进行更新,获得更新后的元学习器
Among them, margin is the set threshold, γ is the balance parameter, and the meta-learner is updated by minimizing the loss function to obtain the updated meta-learner
本发明的有益效果是:本发明针对小样本目标识别场景,在特征级充分挖掘样本的全局特征,在样本级充分探索同类目标不同样本的稳健性特征,在任务级设计元学习器有效累积不同任务的学习经验。通过特征级、样本级和任务级的层级化学习,提升了特征信息的质量, 降低离群样本的负面影响,培养了模型的自主学习能力,进而提高了小样本目标识别技术的鲁棒性。本发明提出的基于层级化元迁移的小样本雷达目标识别方法是一种智能的雷达目标识别方法。The beneficial effects of the present invention are as follows: for small sample target recognition scenarios, the present invention fully mines the global features of samples at the feature level, fully explores the robustness features of different samples of the same target at the sample level, and designs a meta-learner at the task level to effectively accumulate learning experience of different tasks. Through hierarchical learning at the feature level, sample level, and task level, the quality of feature information is improved, the negative impact of outlier samples is reduced, the autonomous learning ability of the model is cultivated, and the robustness of small sample target recognition technology is improved. The small sample radar target recognition method based on hierarchical meta-transfer proposed by the present invention is an intelligent radar target recognition method.
图1为本发明的算法流程图。FIG. 1 is a flow chart of an algorithm of the present invention.
图2为背景技术方法和本发明方法的识别准确率对比图。FIG. 2 is a comparison chart of the recognition accuracy of the background technology method and the method of the present invention.
下面结合附图及实施例,详细描述本发明的技术方案:The technical solution of the present invention is described in detail below in conjunction with the accompanying drawings and embodiments:
如图1所示,本发明设计了一种基于层级化元迁移的小样本雷达目标识别方法,包括特征级,样本级和任务级。针对每个元训练任务,在特征级,采用注意力机制构建特征编码器以提取单个样本中更重要的特征;在样本级,采用注意力机制构建原子编码器,通过整合同类目标不同样本的信息生成高质量的类别原子作为对应类别的代表性信息。在任务级,构建元学习器,通过累积不同元训练任务的学习经验,获得自主学习能力。在面对新的待测任务时,基于少量标记样本,进一步优化训练好的的元学习器,生成高质量的类别原子用于目标识别。将待测样本和类别原子进行对比,选取相似度最高的类别原子的类别作为测试样本的预测类别,完成对测试样本的识别。As shown in Figure 1, the present invention designs a small sample radar target recognition method based on hierarchical meta-transfer, including feature level, sample level and task level. For each meta-training task, at the feature level, an attention mechanism is used to construct a feature encoder to extract more important features in a single sample; at the sample level, an attention mechanism is used to construct an atom encoder, and high-quality category atoms are generated as representative information of the corresponding category by integrating the information of different samples of the same type of target. At the task level, a meta-learner is constructed to acquire autonomous learning ability by accumulating learning experience of different meta-training tasks. When facing a new task to be tested, the trained meta-learner is further optimized based on a small number of labeled samples to generate high-quality category atoms for target recognition. The sample to be tested is compared with the category atom, and the category of the category atom with the highest similarity is selected as the predicted category of the test sample to complete the recognition of the test sample.
实施例:Example:
本例是基于本发明内容方法的实际应用方式,在实际应用中,在建立特征编码器和类别原子编码器的时候进行同步初始化以使其能更快的进行处理。This example is a practical application of the method according to the present invention. In practical applications, synchronous initialization is performed when establishing the feature encoder and the category atom encoder so that they can be processed faster.
步骤1.在源域和目标域中分别采集原始图像样本并进行预处理,初步筛除目标背景的冗余信息,为训练模型做准备。Step 1. Collect and preprocess original image samples in the source domain and target domain respectively, and preliminarily filter out redundant information of the target background to prepare for training the model.
由雷达获取各目标静态时的不同俯仰角下的原始图像,每个固定的俯仰角下,对该目标在不同的方位角下进行观测。将获取的图像根据俯仰角的不同记为源域和目标域,并对其进行切割预处理。The radar obtains the original images of each target at different pitch angles when it is static. At each fixed pitch angle, the target is observed at different azimuth angles. The acquired images are recorded as source domain and target domain according to the different pitch angles, and they are cut and preprocessed.
步骤2.利用样本构建训练任务
每个任务包括支撑集和查询集,用以训练具有自主学习能力的目标识别模型。
Step 2. Use samples to build training tasks Each task includes a support set and a query set to train an object recognition model with autonomous learning capabilities.
记一个K分类任务为
构建所有元训练任务并记为
其中,P是任务总数。针对任务
以K way N shot形式在源域中抽取标记样本构成支撑集并记为
其中,K way N shot指的是对K类目标每个类别随机抽取N个标记训练样本,
是第k类目标的第n个样本;以K way M shot的形式在目标域抽取标记样本构成查询集并记为
其中,
是第k类目标的第m个样本。支撑集和查询集中的样本应是相同类别目标在不同域中的样本,记对应的类别标签为
其中,
Let a K classification task be Construct all meta-training tasks and record them as Where P is the total number of tasks. The support set is composed of labeled samples extracted from the source domain in the form of K way N shot and recorded as Among them, K way N shot means randomly extracting N labeled training samples from each category of K types of targets. is the nth sample of the kth class target; the labeled samples are extracted in the target domain in the form of K way M shot to form a query set and recorded as in, is the mth sample of the kth class target. The samples in the support set and query set should be samples of the same class target in different domains. The corresponding class labels are in,
步骤3、为从不同任务中累积学习经验,培养模型自主学习的能力,通过层级化元迁移模型进行训练学习,对元学习器
进行更新,层级化元迁移模型是由特征级、样本级和任务级构成,具体为:
Step 3: In order to accumulate learning experience from different tasks and cultivate the model's ability to learn autonomously, the meta-learner is trained and learned through the hierarchical meta-transfer model. To update, the hierarchical meta-transfer model is composed of feature level, sample level and task level, specifically:
步骤31.在特征级设计特征编码器
对步骤2中获得的训练任务
的支撑集和查询集分别提取特征,以探索样本深层信息用于识别,进一步地,所述步骤31的具体步骤为:
Step 31. Design feature encoder at feature level For the training task obtained in step 2 The support set and query set are respectively extracted with features to explore the deep information of the sample for identification. Further, the specific steps of step 31 are:
步骤31-1.在特征级设计特征编码器
该特征编码器包含一个神经网络模块和注意力机制模块,神经网络模块具备强大的特征提取能力,可挖掘样本的深层特征,注意力机制模块则是为了使模型有选择地关注样本中的重要信息,提高模型信息处理的效率。利用当前元学习器
中的特征提取器
对其进行初始化:
Step 31-1. Design feature encoder at feature level The feature encoder consists of a neural network module and an attention mechanism module. The neural network module has a strong feature extraction capability and can mine the deep features of the sample. The attention mechanism module is to enable the model to selectively focus on the important information in the sample and improve the efficiency of the model's information processing. Feature extractor in Initialize it:
步骤31-2.采用神经网络模块和注意力机制提取样本的深度全局性特征,具体步骤如下:Step 31-2. Use the neural network module and attention mechanism to extract the deep global features of the sample. The specific steps are as follows:
步骤31-2-1.利用卷积神经网络模块conv(·)对支撑集样本
提取泛化特征,为了清楚的表示,将支撑集样本表示符号简写为S,提取特征过程如下:
Step 31-2-1. Use the convolutional neural network module conv(·) to train the support set samples Extract generalized features. For the sake of clarity, the support set sample representation symbol is abbreviated as S. The feature extraction process is as follows:
步骤31-2-2.将步骤31-2-1得到的样本泛化特征分块并拉直成向量,每个向量的维度是d
1,将所有向量记为[b
1,b
2,…,b
R]
T,其中,R是分块的个数。为了有效整合分块特征中信息,添加一个同维度的可学习向量b
0代表整个样本的全局特征,记嵌入可学习信息的特征为B=[b
0,b
1,b
2,…,b
R]
T,
Step 31-2-2. Divide the sample generalization features obtained in step 31-2-1 into blocks and straighten them into vectors. The dimension of each vector is d 1 . All vectors are recorded as [b 1 , b 2 , …, b R ] T , where R is the number of blocks. In order to effectively integrate the information in the block features, add a learnable vector b 0 of the same dimension to represent the global features of the entire sample. The feature that embeds the learnable information is recorded as B = [b 0 , b 1 , b 2 , …, b R ] T ,
步骤31-2-3.为进一步筛除冗余信息,对步骤31-2-2得到的特征B分别进行变换处理降维到不同的d维嵌入子空间:Step 31-2-3. To further filter out redundant information, the feature B obtained in step 31-2-2 is transformed and reduced to different d-dimensional embedding subspaces:
E=BW
e (2)
E=BW e (2)
U=BW
u (3)
U= BWu (3)
V=BW
v (4)
V=BW v (4)
其中,W
e,W
u,W
v是不同的变换矩阵,E,U,V是不同嵌入子空间中的变换特征,利用注意力机制处理获得全局性特征:
Among them, We , Wu , Wv are different transformation matrices, E, U, V are the transformation features in different embedding subspaces, and the attention mechanism is used to obtain global features:
步骤31-2-4.为缓解梯度消失,对步骤31-2-3得到的全局性特征经过线性映射LN(·)变回d
1维,采用残差结构和步骤31-2-2得到的特征进行结合:
Step 31-2-4. To alleviate the gradient disappearance, the global features obtained in step 31-2-3 are transformed back to d 1 dimensions through linear mapping LN(·), and the residual structure is used to combine with the features obtained in step 31-2-2:
步骤31-3.由于高维空间的信息更加丰富,采用一层全连接网络将步骤31-2得到的特征映射到高维空间,记高维空间的维度是d
2维,再利用一层全连接网络映射回原来的维度d
1维,每层全连接层后面采用激活函数进行处理,以学习获得更加抽象的深层特征
增强信息的表达能力。为避免梯度消失问题,将其和步骤3-2得到的特征采用残差结构进行结合得到深度全局性特征:
Step 31-3. Since the information in high-dimensional space is richer, a layer of fully connected network is used to map the features obtained in step 31-2 to the high-dimensional space. Note that the dimension of the high-dimensional space is d2 , and then a layer of fully connected network is used to map it back to the original dimension d1 . Each fully connected layer is processed with an activation function to learn and obtain more abstract deep features. Enhance the expressiveness of information. To avoid the gradient vanishing problem, combine it with the features obtained in step 3-2 using a residual structure to obtain a deep global feature:
其中,
将对应的可学习向量
取出作为对应样本的深度全局性特征
in, The corresponding learnable vector Take out the deep global features as the corresponding samples
步骤31-4.对任务
的支撑集和查询集进行特征编码:
其中,
分别是任务
支撑集和查询集的深度全局性特征,并且
Step 31-4. Task The support set and query set are feature encoded: in, The tasks are Deep global features of the support and query sets, and
步骤32.在样本级设计基于注意力机制的类别原子编码器
并在当前训练任务
上更新,计算更新后的类别原子,从而提供可靠的代表性信息用于目标识别,进一步地,所述步骤4的具体步骤为:
Step 32. Design an attention-based category atom encoder at the sample level And in the current training task The updated category atoms are calculated to provide reliable representative information for target recognition. Further, the specific steps of step 4 are as follows:
步骤32-1.针对任务
在样本级设计类别原子编码器
并且利用当前元学习器
中的类别原子编码器
对其进行初始化:
Step 32-1. For the task Designing category atom encoders at the sample level And using the current meta-learner Class Atom Encoder in Initialize it:
步骤32-2.利用步骤32-1获得的类别原子编码器
对步骤31获得的支撑集样本的深度全局性特征计算任务
的类别原子,具体步骤如下:
Step 32-2. Use the category atom encoder obtained in step 32-1 The deep global feature calculation task for the support set samples obtained in step 31 The specific steps are as follows:
步骤32-2-1.为去除冗余信息,在不同嵌入子空间中探索样本的深层特征,对支撑集样本深度全局特征
分别进行变换处理降维至d维:
Step 32-2-1. To remove redundant information, explore the deep features of samples in different embedding subspaces and analyze the deep global features of the support set samples. Transform and reduce the dimension to d dimension respectively:
其中,
是不同的变换矩阵,
是不同嵌入子空间中的变换特征,利用注意力机制探索样本级全局性特征:
in, are different transformation matrices, It is the transformation feature in different embedding subspaces, and the attention mechanism is used to explore the sample-level global features:
步骤32-2-2.为缓解梯度消失,对步骤32-2-1得到的样本级全局性特征经过线性映射LN(·)变回d
1维,采用残差结构和步骤31得到的支撑集深度全局性特征进行结合:
Step 32-2-2. To alleviate the gradient vanishing, the sample-level global features obtained in step 32-2-1 are transformed back to d 1 dimensions through linear mapping LN(·), and the residual structure is combined with the support set deep global features obtained in step 31:
步骤32-2-3.由于高维空间的信息更加丰富,采用一层全连接网络将步骤32-2-2得到的特征映射到d
2维的高维空间,再利用一层全连接网络映射回原来的维度d
1维,每层全连接层后面采用激活函数进行处理,以学习获得更加抽象的深层特征
增强信息的表达能力。为避免梯度消失问题,将其和步骤32-2-2得到的特征采用残差结构进行结合得到样本级深度全局性特征:
Step 32-2-3. Since the information in high-dimensional space is richer, a layer of fully connected network is used to map the features obtained in step 32-2-2 to a high-dimensional space of d2 dimensions, and then a layer of fully connected network is used to map it back to the original dimension of d1 dimensions. Each layer of fully connected layer is processed with an activation function to learn and obtain more abstract deep features. Enhance the expressiveness of information. To avoid the gradient vanishing problem, combine it with the features obtained in step 32-2-2 using a residual structure to obtain sample-level deep global features:
步骤32-2-4.对步骤32-2-3得到的样本级深度全局特征取平均,得到样本级注意力机制探索后的类别原子
Step 32-2-4. Average the sample-level deep global features obtained in step 32-2-3 to obtain the category atoms after the sample-level attention mechanism exploration
步骤32-2-5.计算任务
中所有的类别原子并表示为
其中
对应步骤32-2-1至步骤32-2-4的处理流程。
Step 32-2-5. Calculation task All the class atoms in are represented as in Corresponding to the processing flow of step 32-2-1 to step 32-2-4.
步骤32-3.计算步骤31获得的任务
支撑集样本的深度全局性特征和步骤32-2获得的不同类别原子的距离,进一步获得样本
被判定为类别k的概率为:
Step 32-3. Calculate the task obtained in step 31 The deep global features of the support set samples and the distances of different types of atoms obtained in step 32-2 further obtain the samples The probability of being judged as category k is:
其中,dist(·)是距离函数。Here, dist(·) is the distance function.
步骤32-4.根据概率设计并最小化类别原子损失函数,以此更新类别原子编码器和类别原子,具体步骤为:Step 32-4. Design and minimize the category atom loss function according to probability to update the category atom encoder and category atoms. The specific steps are as follows:
步骤32-4-1.设计如下损失函数,使得样本
被判定为类别k的概率尽可能的大,以获取具有识别能力的模型。最小化该损失函数,对类别原子编码器进行更新:
Step 32-4-1. Design the following loss function so that the sample The probability of being judged as category k is as large as possible to obtain a model with recognition ability. Minimize the loss function and update the category atom encoder:
步骤32-4-2.记更新后的模型为
更新后的类别原子为
其中,
Step 32-4-2. Note that the updated model is The updated category atom is in,
步骤33.在任务级累积当前训练任务的学习经验,更新元学习器为
以使元学习器具备自主学习能力来应对新的目标识别任务,进一步地,所述步骤33的具体步骤为:
Step 33. Accumulate the learning experience of the current training task at the task level and update the meta-learner to In order to enable the meta-learner to have autonomous learning capabilities to cope with new target recognition tasks, further, the specific steps of step 33 are:
步骤33-1.计算步骤31获得的任务
查询集样本的深度全局性特征和步骤32获得的不同类别原子的距离,进一步获得样本
被判定为类别k的概率为:
Step 33-1. Calculate the task obtained in step 31 The deep global features of the query set samples and the distances of atoms of different categories obtained in step 32 are further obtained. The probability of being judged as category k is:
其中,dist(·)是距离函数。Here, dist(·) is the distance function.
步骤33-2.根据概率设计元学习器损失函数,最小化该损失函数对元学习器进行更新,获得
具体步骤为:
Step 33-2. Design a meta-learner loss function based on probability, minimize the loss function to update the meta-learner, and obtain The specific steps are:
步骤33-2-1.根据步骤33-1获得的分类概率设计元学习器分类损失函数:Step 33-2-1. Design the meta-learner classification loss function based on the classification probability obtained in step 33-1:
步骤33-2-2.为了提高样本的可分性,提升模型的识别性能,模型训练还采用对比损失 作为损失函数,定义如下:Step 33-2-2. In order to improve the separability of samples and enhance the recognition performance of the model, the model training also uses contrast loss as the loss function, which is defined as follows:
其中,in,
其中,margin是设置的阈值。该约束可以减小样本特征
与对应类别原子
之间的距离,增大与其他类别原子
之间的距离并使其尽可能的大于所设置的阈值。
Among them, margin is the set threshold. This constraint can reduce the sample characteristics With the corresponding class atom The distance between them increases with the distance between atoms of other types. The distance between them should be as large as possible.
步骤33-2-3.将步骤33-2-1和步骤33-2-2的损失函数结合获得总的元学习器损失函数为:Step 33-2-3. Combine the loss functions of step 33-2-1 and step 33-2-2 to obtain the total meta-learner loss function:
其中,γ是平衡参数。最小化元学习器损失函数对元学习器进行更新,获得在任务
上更新后的元学习器
从而累积在任务
的学习经验。
Among them, γ is a balance parameter. Minimize the meta-learner loss function to update the meta-learner and obtain The updated meta-learner Thus accumulating in the task learning experience.
步骤4.更新i=i+1,重复步骤3直至完成所有训练任务上完成多次训练,获得所有元训练任务训练出的元学习器
Step 4. Update i=i+1 and repeat step 3 until all training tasks are completed and multiple trainings are completed to obtain the meta-learner trained by all meta-training tasks.
步骤5.记待测任务的标记样本为支撑集,未标记的待测样本为查询集。利用步骤4获得 的元学习器
对待测样本进行识别,进一步地,所述步骤5的具体步骤为:
Step 5. The labeled samples of the task to be tested are called the support set, and the unlabeled samples to be tested are called the query set. To identify the sample to be tested, further, the specific steps of step 5 are:
步骤5-1.基于在训练任务上累积的学习经验处理待测任务,根据步骤31对待测任务模型进行初始化
并对支撑集和查询集样本提取深度全局性特征。
Step 5-1. Process the task to be tested based on the learning experience accumulated on the training task, and initialize the task model to be tested according to step 31 And extract deep global features for support set and query set samples.
步骤5-2.根据步骤32对待测任务模型进行初始化
利用支撑集计算并更新类别原子;
Step 5-2. Initialize the task model to be tested according to step 32 Calculate and update the category atoms using the support set;
步骤5-3.利用距离函数dist(·)计算查询集待测样本的深度全局性特征和不同类别原子的距离,选取距离最近的类别原子的标签作为待测样本的预测标签,得到识别结果。Step 5-3. Use the distance function dist(·) to calculate the deep global features of the query set sample and the distances between atoms of different categories, select the label of the category atom with the closest distance as the predicted label of the sample to be tested, and obtain the recognition result.
仿真示例Simulation Example
采用实施例模型对运动和静止目标获取与识别MSTAR数据集进行实验,该数据集的传感器采用的是高分辨率的聚束式合成孔径雷达,采用HH极化方式,在X波段工作,分辨率为0.3m×0.3m。该数据大多是静止车辆的SAR切片图像,共包含十类目标,分别是BMP2、T72、BTR70、2S1、BRDM2、BTR60、D7、T62、ZIL131、ZSU234和T72,取其中的7类目标构成元训练任务,剩余的3类目标构建待测任务。以俯仰角17°观测的样本数据为源域样本,俯仰角15°观测的样本数据为目标域样本,实验中的具体样本数目如表1所示。The implementation model is used to experiment with the MSTAR dataset for acquisition and recognition of moving and stationary targets. The sensor of this dataset uses a high-resolution spotlight synthetic aperture radar, adopts HH polarization mode, works in the X-band, and has a resolution of 0.3m×0.3m. Most of the data are SAR slice images of stationary vehicles, which contain a total of ten types of targets, namely BMP2, T72, BTR70, 2S1, BRDM2, BTR60, D7, T62, ZIL131, ZSU234 and T72. Seven types of targets are taken to form the meta-training task, and the remaining three types of targets are used to construct the task to be tested. The sample data observed at a pitch angle of 17° is used as the source domain sample, and the sample data observed at a pitch angle of 15° is used as the target domain sample. The specific number of samples in the experiment is shown in Table 1.
表1实验数据具体数目Table 1 Specific number of experimental data
为去除背景杂波的影响,将样本图像大小以中心切割为64×64。本案例采用的3分类任 务,即每个元训练任务和待测任务均包含三类目标。对于元训练任务,随机选取7类目标中的3类组成元训练任务,针对小样本目标识别,以3way 5shot的形式随机在源域抽取样本构成任务的支撑集,也就是在源域中对该任务的3类目标每类随机抽取5个样本;以3way 15shot的形式在目标域随机抽取样本构成的查询集,也就是对该任务的3类目标每类随机抽取15个样本。对于元训练任务,支撑集和查询集中的样本都是标记样本。以类似的方式对待测目标类别随机抽取样本构成待测任务,其中支撑集来自源域,为俯仰角17°观测的标记样本,查询集来自目标域,为俯仰角15°观测的待测样本。另外,本案例还仿真了不同噪声环境下的目标域样本,对待测任务中查询集的待测样本随机选取一定百分比的像素,并通过用独立的服从均匀分布的样本替换其像素点的强度来破坏像素,添加的随机噪声服从[0,μ
max]的均匀分布,其中,μ
max是图像中像素点中的最大值。选取的像素比例分别为0%,5%和15%,分别代表不同噪声环境下的目标域,其中,0%表示原有数据集中的俯仰角15°观测样本构建的待测样本。
To remove the influence of background clutter, the sample image size is cut into 64×64 at the center. This case uses a 3-classification task, that is, each meta-training task and the task to be tested contains three types of targets. For the meta-training task, 3 of the 7 types of targets are randomly selected to form the meta-training task. For small sample target recognition, samples are randomly extracted from the source domain in the form of 3way 5shot to form the support set of the task, that is, 5 samples are randomly extracted from each of the 3 types of targets in the source domain for this task; the query set is composed of samples randomly extracted from the target domain in the form of 3way 15shot, that is, 15 samples are randomly extracted from each of the 3 types of targets in this task. For the meta-training task, the samples in the support set and the query set are all labeled samples. In a similar way, samples of the target category to be tested are randomly extracted to form the task to be tested, where the support set comes from the source domain and is the labeled samples observed at a pitch angle of 17°, and the query set comes from the target domain and is the samples to be tested observed at a pitch angle of 15°. In addition, this case also simulates target domain samples under different noise environments. A certain percentage of pixels are randomly selected from the test samples of the query set in the test task, and the pixels are destroyed by replacing the intensity of their pixels with independent samples that obey the uniform distribution. The added random noise obeys the uniform distribution of [0, μ max ], where μ max is the maximum value of the pixels in the image. The selected pixel ratios are 0%, 5%, and 15%, respectively, representing the target domains under different noise environments, where 0% represents the test samples constructed from the 15° pitch angle observation samples in the original data set.
本发明针对小样本目标识别设计了不同噪声环境下的实验来验证提出算法的优越性,分别对比了背景技术方法和本发明方法对待测任务的识别结果。实验中特征编码器的神经网络模块由四层卷积层组成,每层卷积层后均采用最大池化运算来缩减模型的大小,提高计算速度,表2展示了每层卷积层及池化运算的详细参数,包括卷积核的尺寸,卷积时的步长,填充的尺寸以及池化窗的尺寸。此外,实验中其他参数具体设置为:R=3,d
1=252,d=64,d
2=128,γ=0.01,以及margin=200。采用200个元训练任务进行训练,采用1000个待测任务的平均识别率为算法性能的量化指标。随着目标域噪声水平的增大,背景技术方法均产生了不同程度的明显下降,其中,背景技术方法1在0%和15%的噪声环境下的识别准确率分别为77.43%和71.66%,背景技术方法的识别准确率为71.67%和68.1%,而本发明方法仍能保持较高的识别率,在0%,5%和15%噪声环境下的识别准确率分别为83.86%,82.24%,和81.92%,具有明显优势。综上所述,实验结果证明了本发明有效地在小样本目标识别场景下探索了样本的深度全局性特征,培养了模型的自主学习能力,建立了更加稳定的元学习模型,提升了目标识别性能。
The present invention designs experiments in different noise environments for small sample target recognition to verify the superiority of the proposed algorithm, and compares the recognition results of the background technology method and the method of the present invention on the task to be tested. In the experiment, the neural network module of the feature encoder consists of four convolutional layers, and the maximum pooling operation is used after each convolutional layer to reduce the size of the model and improve the calculation speed. Table 2 shows the detailed parameters of each convolutional layer and pooling operation, including the size of the convolution kernel, the step size during convolution, the padding size, and the size of the pooling window. In addition, other parameters in the experiment are specifically set to: R = 3, d 1 = 252, d = 64, d 2 = 128, γ = 0.01, and margin = 200. 200 meta-training tasks are used for training, and the average recognition rate of 1000 tasks to be tested is used as a quantitative indicator of the algorithm performance. As the noise level in the target domain increases, the background technology methods all show a significant decline to varying degrees. Among them, the recognition accuracy of background technology method 1 in 0% and 15% noise environments is 77.43% and 71.66% respectively, and the recognition accuracy of the background technology method is 71.67% and 68.1%, while the method of the present invention can still maintain a high recognition rate, with recognition accuracy rates of 83.86%, 82.24%, and 81.92% in 0%, 5%, and 15% noise environments, respectively, which has obvious advantages. In summary, the experimental results prove that the present invention effectively explores the deep global features of samples in small sample target recognition scenarios, cultivates the autonomous learning ability of the model, establishes a more stable meta-learning model, and improves target recognition performance.
表2实验参数设置Table 2 Experimental parameter settings
卷积层Convolutional Layer | 卷积核尺寸Convolution kernel size | 步长Step Length | 填充尺寸Filling size | 池化窗尺寸Pooling window size |
第一层level one | 5×55×5 | 11 | 00 | 2×22×2 |
第二层Second floor | 3×33×3 | 11 | 00 | 2×22×2 |
第三层the third floor | 3×33×3 | 11 | 11 | 2×22×2 |
第四层Fourth floor | 3×33×3 | 11 | 11 | 2×22×2 |
Claims (5)
- 一种基于层级化元迁移的小样本雷达目标识别方法,其特征在于,包括以下步骤:A small sample radar target recognition method based on hierarchical element migration is characterized by comprising the following steps:S1、通过雷达获取各目标静态时在源域和目标域的原始图像,将对目标在不同方位角下进行观测得到的图像进行切割处理后得到样本;S1. Obtaining original images of each target in the source domain and the target domain when the target is static through radar, and cutting the images obtained by observing the target at different azimuth angles to obtain samples;S2、利用样本构建训练任务 其中P是任务总数,任务 包括支撑集和查询集,其中支撑集是从源域中抽取标记样本构成,查询集是从目标域中抽取标记样本构成; S2. Use samples to build training tasks Where P is the total number of tasks, It includes support set and query set, where the support set is composed of labeled samples extracted from the source domain, and the query set is composed of labeled samples extracted from the target domain;S3、通过层级化元迁移模型进行训练学习,对元学习器 进行更新,具体为: S3. Training and learning through hierarchical meta-transfer models Update, specifically:S31、在特征级构建基于注意力机制的特征编码器 利用元学习器对特征编码器进行初始化 后,提取 中支撑集和查询集的深度全局性特征; S31. Constructing a feature encoder based on the attention mechanism at the feature level Initialize the feature encoder using a meta-learner After that, extract Deep global features of the support set and query set;S32、在样本级构建基于注意力机制的类别原子编码器 利用元学习器对类别原子编码器进行初始化 后,基于获得的 支撑集样本的深度全局性特征计算 的类别原子,根据支撑集样本和不同类别原子的距离获得对应样本归属于不同类的概率,再根据概率设计并最小化类别原子损失函数,以此更新类别原子编码器和类别原子; S32. Constructing a category atom encoder based on attention mechanism at sample level Initialize the category atom encoder using a meta-learner Afterwards, based on the obtained Deep global feature calculation of support set samples The category atoms of the support set are obtained, and the probability of the corresponding samples belonging to different categories is obtained according to the distance between the support set samples and the different category atoms. Then, the category atom loss function is designed and minimized according to the probability to update the category atom encoder and category atoms.S33、在任务级累计当前训练任务的学习经验,更新元学习器:S33. Accumulate the learning experience of the current training task at the task level and update the meta-learner:根据 查询集样本的深度全局性特征和不同类别原子的距离,获得对应样本归属于不同类的概率,根据概率设计元学习器损失函数,最小化该损失函数对元学习器进行更新,获得更新后的元学习器 according to The deep global features of the query set samples and the distances between atoms of different categories are used to obtain the probability that the corresponding samples belong to different categories. The meta-learner loss function is designed based on the probability, and the meta-learner is updated by minimizing the loss function to obtain the updated meta-learner.S4、通过重复步骤S3完成所有训练任务,获得所有元训练任务训练出的元学习器,记训 练出的元学习器为 S4. Complete all training tasks by repeating step S3 to obtain the meta-learner trained by all meta-training tasks. The trained meta-learner is recorded asS5、记待测任务的标记样本为支撑集,未标记的待测样本为查询集;利用S4获得的元学习器进行初始化 得到目标识别用特征编码器和类别原子编码器,利用目标识别用特征编码对支撑集和查询集样本提取深度全局性特征;利用目标识别用类别原子编码器基于支撑集深度全局性特征计算并更新类别原子,利用距离函数dist(·)计算查询集中待测样本的深度全局性特征和不同类别原子的距离,选取距离最近的类别原子的标签作为待测样本的预测标签,得到识别结果。 S5: The labeled samples of the task to be tested are the support set, and the unlabeled samples to be tested are the query set; the meta-learner obtained in S4 is used for initialization A feature encoder for target recognition and a category atom encoder are obtained, and the feature encoder for target recognition is used to extract deep global features for the support set and query set samples. The category atom encoder for target recognition is used to calculate and update the category atoms based on the deep global features of the support set, and the distance function dist(·) is used to calculate the distance between the deep global features of the sample to be tested in the query set and the atoms of different categories, and the label of the category atom with the closest distance is selected as the predicted label of the sample to be tested to obtain the recognition result.
- 根据权利要求1所述的一种基于层级化元迁移的小样本雷达目标识别方法,其特征在于,步骤S2中,所述支撑集是通过K way N shot形式在源域中抽取标记样本构成,定义为 K way N shot指的是对K类目标每个类别随机抽取N个标记训练样本, 是第k类目标的第n个样本;查询集是通过K way N shot的形式在目标域抽取标记样本构成,定义为 其中, 是第k类目标的第m个样本;支撑集和查询集中的样本是相同类别目标在不同域中的样本,定义对应的类别标签为 其中, The small sample radar target recognition method based on hierarchical meta-transfer according to claim 1 is characterized in that in step S2, the support set is formed by extracting labeled samples in the source domain in the form of K way N shot, defined as K way N shot means randomly extracting N labeled training samples from each category of K target. is the nth sample of the kth class target; the query set is composed of labeled samples extracted in the target domain in the form of K way N shot, defined as in, is the mth sample of the kth class target; the samples in the support set and query set are samples of the same class target in different domains, and the corresponding class labels are defined as in,
- 根据权利要求2所述的一种基于层级化元迁移的小样本雷达目标识别方法,其特征在于,步骤S31中,所述特征编码器 包括一个神经网络模块和注意力机制模块,提取深度全局性特征的具体方式为: The small sample radar target recognition method based on hierarchical element migration according to claim 2 is characterized in that in step S31, the feature encoder It includes a neural network module and an attention mechanism module. The specific method of extracting deep global features is as follows:通过神经网络模块对样本提取泛化特征;Extract generalized features from samples through neural network modules;将泛化特征分块并拉直成向量,每个向量的维度是d 1,记为 [b 1,b 2,…,b R] T,其中,R是分块的个数,添加一个同维度的可学习向量b 0代表整个样本的全局特征,记嵌入可学习信息后的特征为 The generalized features are divided into blocks and straightened into vectors. The dimension of each vector is d 1 , denoted as [b 1 , b 2 , …, b R ] T , where R is the number of blocks. A learnable vector b 0 of the same dimension is added to represent the global features of the entire sample. The features after embedding the learnable information are denoted as将特征B分别进行变换处理降维到不同的d维嵌入子空间:Transform feature B and reduce its dimension to different d-dimensional embedding subspaces:E=BW e E=BW eU=BW u U= BWuV=BW v V=BW v其中,W e,W u,W v是不同的变换矩阵,E,U,V是不同嵌入子空间中的变换特征,利用注意力机制处理获得全局性特征 Among them, We , Wu , Wv are different transformation matrices, E, U, V are the transformation features in different embedding subspaces, and the attention mechanism is used to obtain global features.将全局性特征经过线性映射LN(·)变回d 1维,采用残差结构和特征B进行结合得到 The global features are transformed back to d1 dimensions through linear mapping LN(·), and the residual structure is combined with feature B to obtain通过全连接层将特征B先映射到高维空间,记高维空间的维数为d 2维,再映射回d 1维的低维空间,得到深层特征 与特征 采用残差结构进行结合得到深度全局性特征 将可学习向量 取出作为对应样本的深度全局性特征 Feature B is first mapped to a high-dimensional space through a fully connected layer, and the dimension of the high-dimensional space is recorded as d2 , and then mapped back to a low-dimensional space of d1 to obtain the deep feature With features Use residual structure to combine and obtain deep global features The learnable vector Take out the deep global features as the corresponding samples
- 根据权利要求3所述的一种基于层级化元迁移的小样本雷达目标识别方法,其特征在于,步骤S32中,所述更新类别原子编码器和类别原子的具体方法为:The small sample radar target recognition method based on hierarchical element migration according to claim 3 is characterized in that in step S32, the specific method of updating the category atom encoder and the category atom is:利用S31中提取的支撑集深度全局性特征 将 分别进行变换处理降维到不同的d维嵌入子空间: Using the deep global features of the support set extracted from S31 Will Transform and reduce the dimensions to different d-dimensional embedding subspaces:其中, 和 是不同的变换矩阵, 和 是不同嵌入子空间中的变换特征,利用注意力机制得到样本级全局性特征 in, and are different transformation matrices, and It is the transformation feature in different embedding subspaces, and the sample-level global feature is obtained by using the attention mechanism将样本级全局性特征经过线性映射LN(·)变回d 1维,采用残差结构和深度全局性特征进行结合得到 The sample-level global features are transformed back to d1 dimensions through linear mapping LN(·), and the residual structure and deep global features are combined to obtain通过全连接层先将特征 映射到d 2维的高维空间,再映射回d 1维的低维空间,得到深层特征 与特征 采用残差结构进行结合得到样本级深度全局性特征 Through the fully connected layer, the features are first Map to a high-dimensional space of d2 , and then map back to a low-dimensional space of d1 to obtain deep features With features The residual structure is used to combine and obtain sample-level deep global features对样本级深度全局特征取平均,得到样本级类别原子 The sample-level deep global features are averaged to obtain the sample-level category atoms利用类别原子编码器基于得到的支撑集样本的深度全局性特征计算任务 中所有的类别原子并表示为 其中 Using the category atomic encoder to calculate the deep global features of the support set samples All the class atoms in are represented as in根据获得的任务 支撑集样本的深度全局性特征和获得的不同类别原子的距离,获得样本 被判定为类别k的概率为: According to the tasks obtained The deep global features of the support set samples and the distances of atoms of different categories obtained to obtain the samples The probability of being judged as category k is:其中,dist(·)是距离函数;Where dist(·) is the distance function;根据概率设计并最小化类别原子损失函数:Design and minimize the category atom loss function based on probability:
- 根据权利要求4所述的一种基于层级化元迁移的小样本雷达目标识别方法,其特征在于,步骤S33中,更新元学习器的具体方法为:The small sample radar target recognition method based on hierarchical meta-transfer according to claim 4 is characterized in that in step S33, the specific method of updating the meta-learner is:根据获得的任务 查询集样本的深度全局性特征和获得的不同类别原子的距离,获得样本 被判定为类别k的概率为: According to the tasks obtained The deep global features of the query set samples and the distances of atoms of different categories are obtained to obtain the samples The probability of being judged as category k is:根据概率设计元学习器损失函数:Design the meta-learner loss function based on probability:
Applications Claiming Priority (2)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202211276348.4 | 2022-10-19 | ||
CN202211276348.4A CN115345322B (en) | 2022-10-19 | 2022-10-19 | Small sample radar target identification method based on hierarchical element migration |
Publications (1)
Publication Number | Publication Date |
---|---|
WO2024082374A1 true WO2024082374A1 (en) | 2024-04-25 |
Family
ID=83957489
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
PCT/CN2022/133980 WO2024082374A1 (en) | 2022-10-19 | 2022-11-24 | Few-shot radar target recognition method based on hierarchical meta transfer |
Country Status (2)
Country | Link |
---|---|
CN (1) | CN115345322B (en) |
WO (1) | WO2024082374A1 (en) |
Families Citing this family (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN115345322B (en) * | 2022-10-19 | 2023-02-07 | 电子科技大学长三角研究院(衢州) | Small sample radar target identification method based on hierarchical element migration |
Citations (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20210003700A1 (en) * | 2019-07-02 | 2021-01-07 | Wuyi University | Method and apparatus for enhancing semantic features of sar image oriented small set of samples |
CN112990334A (en) * | 2021-03-29 | 2021-06-18 | 西安电子科技大学 | Small sample SAR image target identification method based on improved prototype network |
CN114488140A (en) * | 2022-01-24 | 2022-05-13 | 电子科技大学 | Small sample radar one-dimensional image target identification method based on deep migration learning |
CN114492581A (en) * | 2021-12-27 | 2022-05-13 | 中国矿业大学 | Method for classifying small sample pictures based on transfer learning and attention mechanism element learning application |
CN114511739A (en) * | 2022-01-25 | 2022-05-17 | 哈尔滨工程大学 | Task-adaptive small sample image classification method based on meta-migration learning |
CN115345322A (en) * | 2022-10-19 | 2022-11-15 | 电子科技大学长三角研究院(衢州) | Small sample radar target identification method based on hierarchical element migration |
Family Cites Families (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN114387524B (en) * | 2022-03-24 | 2022-06-03 | 军事科学院系统工程研究院网络信息研究所 | Image identification method and system for small sample learning based on multilevel second-order representation |
CN114859316A (en) * | 2022-06-14 | 2022-08-05 | 中国人民解放军海军航空大学 | Radar target intelligent identification method based on task relevance weighting |
CN114879185A (en) * | 2022-06-14 | 2022-08-09 | 中国人民解放军海军航空大学 | Radar target intelligent identification method based on task experience migration |
-
2022
- 2022-10-19 CN CN202211276348.4A patent/CN115345322B/en active Active
- 2022-11-24 WO PCT/CN2022/133980 patent/WO2024082374A1/en unknown
Patent Citations (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20210003700A1 (en) * | 2019-07-02 | 2021-01-07 | Wuyi University | Method and apparatus for enhancing semantic features of sar image oriented small set of samples |
CN112990334A (en) * | 2021-03-29 | 2021-06-18 | 西安电子科技大学 | Small sample SAR image target identification method based on improved prototype network |
CN114492581A (en) * | 2021-12-27 | 2022-05-13 | 中国矿业大学 | Method for classifying small sample pictures based on transfer learning and attention mechanism element learning application |
CN114488140A (en) * | 2022-01-24 | 2022-05-13 | 电子科技大学 | Small sample radar one-dimensional image target identification method based on deep migration learning |
CN114511739A (en) * | 2022-01-25 | 2022-05-17 | 哈尔滨工程大学 | Task-adaptive small sample image classification method based on meta-migration learning |
CN115345322A (en) * | 2022-10-19 | 2022-11-15 | 电子科技大学长三角研究院(衢州) | Small sample radar target identification method based on hierarchical element migration |
Also Published As
Publication number | Publication date |
---|---|
CN115345322B (en) | 2023-02-07 |
CN115345322A (en) | 2022-11-15 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN111583263B (en) | Point cloud segmentation method based on joint dynamic graph convolution | |
CN110956185A (en) | Method for detecting image salient object | |
CN112446423B (en) | Fast hybrid high-order attention domain confrontation network method based on transfer learning | |
CN112347970B (en) | Remote sensing image ground object identification method based on graph convolution neural network | |
Wang et al. | Distilling knowledge from an ensemble of convolutional neural networks for seismic fault detection | |
CN112990334A (en) | Small sample SAR image target identification method based on improved prototype network | |
CN111079847B (en) | Remote sensing image automatic labeling method based on deep learning | |
WO2022218396A1 (en) | Image processing method and apparatus, and computer readable storage medium | |
CN113221848B (en) | Hyperspectral open set field self-adaption method based on multi-classifier domain confrontation network | |
WO2024082374A1 (en) | Few-shot radar target recognition method based on hierarchical meta transfer | |
He et al. | SAR target recognition and unsupervised detection based on convolutional neural network | |
CN110263855A (en) | A method of it is projected using cobasis capsule and carries out image classification | |
CN115292532A (en) | Remote sensing image domain adaptive retrieval method based on pseudo label consistency learning | |
CN113420593B (en) | Small sample SAR automatic target recognition method based on hybrid inference network | |
Zhang et al. | SaltISCG: Interactive salt segmentation method based on CNN and graph cut | |
Gu et al. | A classification method for polsar images using SLIC superpixel segmentation and deep convolution neural network | |
CN111209813B (en) | Remote sensing image semantic segmentation method based on transfer learning | |
CN112270285A (en) | SAR image change detection method based on sparse representation and capsule network | |
CN113409351B (en) | Unsupervised field self-adaptive remote sensing image segmentation method based on optimal transmission | |
Ma et al. | Region of interest extraction based on unsupervised cross-domain adaptation for remote sensing images | |
CN113239895A (en) | SAR image change detection method of capsule network based on attention mechanism | |
Bao et al. | Contextual vision transformers for robust representation learning | |
CN113627480A (en) | Polarized SAR image classification method based on reinforcement learning | |
Shkodrani et al. | Dynamic adaptation on non-stationary visual domains | |
Yang et al. | Agricultural Few-Shot Selection by Model Confidences for Multimedia Internet of Things Acquisition Dataset |