Traditional supervised learning methods have achieved great success in automatic target recognition (ATR). Unfortunately, if the model is trained exclusively on new class samples, the model will forget all knowledge about the old class samples. This phenomenon is called catastrophic forgetting. Incremental learning method can prevent catastrophic forgetting by keeping a small number of old class samples as exemplars and training them together with new class samples. However, the ratio of old to new class samples is still seriously out of balance. In this paper, an data augmentation method of old class samples is proposed by using random interpolation (RI) between two exemplars to generate fake samples in the process of incremental learning. In this method, the distribution of the original classes is partially restored while also creating a numerical balance between the old and new class samples. Experiments on the Moving and Stationary Target Acquisition and Recognition (MSTAR) dataset demonstrate the effectiveness of this method in Incremental SAR ATR.

To enable the synthetic aperture radar (SAR) automatic target recognition (ATR) system to continuously adapt to new recognition scenarios, it is necessary to equip the system with the ability to quickly update models. However, when these models learn new tasks, the knowledge of old tasks is quickly forgotten, a phenomenon known as catastrophic forgetting. The reason for catastrophic forgetting is that the model does not use the features of old tasks sufficiently. In this letter, an exemplar-free class incremental learning based on maximizing mutual information (CIL-MMI) is proposed to solve this problem. To effectively use the extracted features, CIL-MMI actively clusters features to maximize the mutual information (MI) between features and corresponding labels. The proposed method successfully avoids the distribution overlap caused by the small interclass differences and large intraclass variances inherent in SAR images. Experiments on the Moving and Stationary Target Acquisition and Recognition (MSTAR) dataset indicate that the proposed method outperforms state-of-the-art approaches, demonstrating improvements of 5.41%, 1.93%, and 2.47% at incremental steps 1, 2, and 3, respectively.

The availability of a large amount of labeled data has promoted the success of deep learning in Synthetic aperture radar (SAR) target recognition tasks, but these labeled data are invariably obtained manually, which requires huge labor costs. With the development of the SAR field, SAR target categories are still increasing, so there is a large amount of data to be marked. In this paper, we use labeled data to discover novel categories, thereby reducing the cumbersome labeling process. We innovatively propose to apply AutoMix to the labeled dataset to expand the neighborhood distribution and motivate the discrete sample space continuously, which greatly improves the generalization ability of the model. Three stage learning is used to solve the problem of novel class discovery. Firstly, self-supervised learning is used to learn common features from labeled data and unlabeled data. Secondly, AutoMix is used to labeled data which is used to train the feature extractor by supervised learning. Finally, the knowledge of labeled data is transferred to unlabeled images to generate pairs of pseudo-labels for clustering. We experimentally proved that our model can effectively discover novel categories in unlabeled data.

The traditional Synthetic Aperture Radar Automatic Target Recognition (SAR/ATR) algorithm can train a sufficient number of known class samples and classify the samples in the test set. However, if the old model is trained only with the new class samples, the old class samples’ knowledge is easily forgotten by the new model, which is called catastrophic forgetting. The reason is that the model only fits the distribution of current training samples, so training the whole data set is necessary. Due to the limitation of storage resources, it is often not feasible to retain the whole data set. In order to avoid this phenomenon, a small number of old class samples can be kept to train with the new class samples. Therefore, how to select the old class samples becomes the key point. In this paper, the Density Coverage-Based Exemplar Selection (DCBES) is proposed to choose the key samples of the old class. DCBES selects samples based on the metric learning theory and the set covering theory. First, the metric learning theory is used to measure the similarity between samples and to obtain the density range of samples. Then the exemplar selection problem is considered a set covering problem, to select a fixed number of exemplars to achieve the maximum coverage of the class density range. Experiments on the Moving and Stationary Target Acquisition and Recognition (MSTAR) dataset show that our method is superior to other exemplar selection methods and achieves the best results.

Although deep learning has achieved great success in automatic target recognition, the model needs a large number of labeled samples for training. In real life, it is a time-consuming and laborious work to label unlabeled samples, so how to find unknown classes from a large number of unlabeled samples has aroused widespread concern. In this paper, we study how to discover unseen class from an unlabeled image set under the assumption that there are samples related to the unseen class but of different classes as prior knowledge. The Automatic Unseen Class Discovery (AUCD) algorithm is proposed in this paper, which mainly solves the problem of unseen class discovery from two aspects, one is how to actively form clusters according to their classes for unknown class samples, and the other is how to obtain the number of formed clusters. Several experiments based on the Moving and Stationary Target Acquisition and Recognition (MSTAR) dataset prove the effectiveness of the proposed approach in the field of unseen class discovery.

Although deep learning methods have achieved great success in synthetic aperture radar automatic target recognition (SAR ATR), their accuracies decline sharply as new classes are learned, which is known as catastrophic forgetting. The overlapping or confusion between the representations of new and old classes in the feature space is the main cause of catastrophic forgetting. In this paper, the Incremental Class Anchor Clustering (ICAC) is proposed to address this issue. ICAC solves this problem from three perspectives: first, how to learn the new classes; second, how to enable the model to recognize and classify the old classes; third, how to solve the imbalance between old classes and new classes. To learn the new classes, ICAC adaptively adds new anchored class centers for new classes, and the features of each new class will be clustered around the corresponding anchored class center. To enable the model to recognize and classify the old classes, ICAC stores some exemplars for the old classes to ensure the classification ability of the old classes without losing the old class centers in the feature space. At the same time, ICAC adopts knowledge distillation to further alleviate catastrophic forgetting. To solve the imbalance between old classes and new classes, ICAC proposes a learning strategy named Separable Learning (SL), which computes the losses of the old and new exemplars separately and then adds the two losses to make a gradient descent. Experiments on the MSTAR dataset and OpenSARShip dataset demonstrate the effectiveness of this method in SAR automatic targets recognition.