With the evolutionary development of modern communications technology, automatic modulation classification (AMC) has played an increasing role in the complex wireless communication environment. Existing AMC schemes based on deep learning use a neural network to extract features and calculate feature maps, then feed them into fully connected layers for classification. However, existing schemes still are insufficient in utilizing feature maps. To overcome this limitation, a novel adaptive wavelet network (AWN) is proposed, which combines adaptive wavelet decomposition based on the lifting scheme and channel attention mechanism.In contrast to the previous models, the multi-level decomposition of AWN explicitly extracts the features of multiple frequency bands. The channel attention mechanism further selects the optimal frequencies from the candidate frequencies. AWN explores a novel AMC paradigm that efficiently integrates the inherent properties of the signal by introducing prior knowledge in the frequency domain. Simulation results demonstrate that our proposed AMC scheme outperforms the benchmark scheme and has rather low computational complexity.

Internet of Things (IoT) networks are often subject to many malicious attacks in untrusted environments, and automatic modulation classification (AMC) is an effective way to combat IoT physical-layer threats. However, most existing AMC methods assume sufficient labeled signals and invariant signal distribution, which is often impossible in untrusted environments. In this article, a new meta-learning method is proposed for a few-shot AMC with distribution bias. First, a multi-frequency octave ResNet (MFOR) is constructed to learn coarse (low-frequency) and fine (high-frequency) features, which can efficiently identify the modulation type of the signal while saving computational resources. Second, a large number of classification-related meta-tasks are established for training MFOR to explore general knowledge in signal classification, and then transfer it to the AMC. Different with deep neural networks (DNNs) that learn a mapping by multiple instances, the MFOR with meta-learning (denoted as M-MFOR) can improve the generalization ability of new AMC tasks with very few instances and distribution bias. Furthermore, we find that the distribution bias between data can be reduced by adjusting the normalized distribution and propose a class-related mixup. Extensive experiments are taken on several datasets to investigate the effectiveness of M-MFOR. The results show its feasibility and superiority over existing methods.

In this paper, an Adaptive Fusion Transformer (AFT) is proposed for unsupervised pixel-level fusion of visible and infrared images. Different from the existing convolutional networks, transformer is adopted to model the relationship of multi-modality images and explore cross-modal interactions in AFT. The encoder of AFT uses a Multi-Head Self-attention (MSA) module and Feed Forward (FF) network for feature extraction. Then, a Multi-head Self-Fusion (MSF) module is designed for the adaptive perceptual fusion of the features. By sequentially stacking the MSF, MSA, and FF, a fusion decoder is constructed to gradually locate complementary features for recovering informative images. In addition, a structure-preserving loss is defined to enhance the visual quality of fused images. Extensive experiments are conducted on several datasets to compare our proposed AFT method with 21 popular approaches. The results show that AFT has state-of-the-art performance in both quantitative metrics and visual perception.

Automatic Modulation Classification (AMC) is a critical task in industrial cognitive communication systems. Existing state-of-the-art methods, typified by real-valued convolutional neural networks, have introduced innovative solutions for AMC. However, such models viewed the two constituent components of complex-valued modulated signals as discrete real-valued inputs, causing structural phase damage to original signals and reduced interpretability of the model. In this article, a novel end-to-end AMC model called a complex-valued depth-wise separable convolutional neural network (CDSCNN) is proposed, which adopts complex-valued operation units to enable automatic complex-valued feature learning specifically tailored for AMC. Considering the limited hardware resources available in industrial scenarios, complex-valued depth-wise separable convolution (CDSC) is designed to strike a balance between classification accuracy and model complexity. With an overall accuracy of 62.63% on the RadioML2016.10a dataset, CDSCNN outperforms its counterparts by 1% to 11%. After fine-tuning on the RadioML2016.10b dataset, the overall accuracy reaches 63.15%, demonstrating the robust recognition and generalization capability of CDSCNN. Moreover, the CDSCNN exhibits lower model complexity compared to other methods.

Recently, the multiscale problem in computer vision has gradually attracted people’s attention. This article focuses on multiscale representation for object detection and recognition, comprehensively introduces the development of multiscale deep learning, and constructs an easy-to-understand, but powerful knowledge structure. First, we give the definition of scale, explain the multiscale mechanism of human vision, and then lead to the multiscale problem discussed in computer vision. Second, advanced multiscale representation methods are introduced, including pyramid representation, scale-space representation, and multiscale geometric representation. Third, the theory of multiscale deep learning is presented, which mainly discusses the multiscale modeling in convolutional neural networks (CNNs) and Vision Transformers (ViTs). Fourth, we compare the performance of multiple multiscale methods on different tasks, illustrating the effectiveness of different multiscale structural designs. Finally, based on the in-depth understanding of the existing methods, we point out several open issues and future directions for multiscale deep learning.

Semantic segmentation based on deep learning has achieved impressive results in recent years, but these results are supported by a large amount of labeled data which requires intensive annotation at the pixel level, particularly for high-resolution remote sensing (RS) images. In this work, we propose a simple yet efficient semisupervised learning framework based on linear sampling self-training, named LSST, to improve the performance of RS image semantic segmentation. Specifically, the classical pseudo-labeling-based self-training paradigm is enhanced by injecting strong data augmentations (SDA) applicable to RS images, based on which a powerful baseline is constructed. Nevertheless, the problem of insufficient data training to generate pseudo-labels with a high level of noise persists, and the noisy pseudo-labels will continue to accumulate and impede model improvement during the re-training phase. Previous works commonly employ a pre-defined threshold to remove noise, but it will lead to overfitting the model to easily identified classes. To address it, a method using linear sampling (LS) is presented for assigning thresholds to different classes in an adaptive manner, which provides noiseless regions for re-training. Experiments prove that the proposed pixel-wise selection is more available for segmentation than image-level selection in RS images. Finally, LSST achieves state-of-the-art on several datasets and different evaluation metrics. The source code of the this paper is available at https://github.com/xiaoqiang-lu/LSST.