Search

Convolutional neural network (CNN)-based and Transformer-based methods for hyperspectral image (HSI) classification have rapidly advanced due to their unique characterization capabilities. However, the fixed kernel sizes in convolutional layers limit the comprehensive utilization of multi-scale features in HSI land cover analysis, while the Transformer's multi-head self-attention (MHSA) mechanism faces challenges in effectively encoding feature information across various dimensions. To tackle this issue, this article introduces an HSI classification method, based on multi-scale convolutional features and multi-attention mechanisms (i.e., MSCF-MAM). Firstly, the model employs a multi-scale convolutional module to capture features across different scales in HSIs. Secondly, to enhance the integration of local and global channel features and establish long-range dependencies, a feature enhancement module based on pyramid squeeze attention (PSA) is employed. Lastly, the model leverages a classical Transformer Encoder (TE) and linear layers to encode and classify the transformed spatial–spectral features. The proposed method is evaluated on three publicly available datasets—Salina Valley (SV), WHU-Hi-HanChuan (HC), and WHU-Hi-HongHu (HH). Extensive experimental results have demonstrated that the MSCF-MAM method outperforms several representative methods in terms of classification performance. [ABSTRACT FROM AUTHOR]

Hyperspectral image classification (HSIC) has garnered increasing attention among researchers. While classical networks like convolution neural networks (CNNs) have achieved satisfactory results with the advent of deep learning, they are confined to processing local information. Vision transformers, despite being effective at establishing long-distance dependencies, face challenges in extracting high-representation features for high-dimensional images. In this paper, we present the multiscale efficient attention with enhanced feature transformer (MEA-EFFormer), which is designed for the efficient extraction of spectral–spatial features, leading to effective classification. MEA-EFFormer employs a multiscale efficient attention feature extraction module to initially extract 3D convolution features and applies effective channel attention to refine spectral information. Following this, 2D convolution features are extracted and integrated with local binary pattern (LBP) spatial information to augment their representation. Then, the processed features are fed into a spectral–spatial enhancement attention (SSEA) module that facilitates interactive enhancement of spectral–spatial information across the three dimensions. Finally, these features undergo classification through a transformer encoder. We evaluate MEA-EFFormer against several state-of-the-art methods on three datasets and demonstrate its outstanding HSIC performance. [ABSTRACT FROM AUTHOR]

Background: In advanced Parkinson's disease (PD), axial symptoms are common and can be debilitating. Although deep brain stimulation (DBS) significantly improves motor symptoms, conventional high-frequency stimulation (HFS) has limited effectiveness in improving axial symptoms. In this study, we investigated the effects on multiple axial symptoms after DBS surgery with three different frequency programming paradigms comprising HFS, low-frequency stimulation (LFS), and variable-frequency stimulation (VFS). Methods: This study involved PD patients who had significant preoperative axial symptoms and underwent bilateral subthalamic nucleus (STN) DBS. Axial symptoms, motor symptoms, medications, and quality of life were evaluated preoperatively (baseline). One month after surgery, HFS was applied. At 6 months post-surgery, HFS assessments were performed, and HFS was switched to LFS. A further month later, we conducted LFS assessments and switched LFS to VFS. At 8 months after surgery, VFS assessments were performed. Results: Of the 21 PD patients initially enrolled, 16 patients were ultimately included in this study. Regarding HFS, all axial symptoms except for the Berg Balance Scale (p < 0.0001) did not improve compared with the baseline (all p > 0.05). As for LFS and VFS, all axial symptoms improved significantly compared with both the baseline and HFS (all p < 0.05). Moreover, motor symptoms and medications were significantly better than the baseline (all p < 0.05) after using LFS and VFS. Additionally, the quality of life of the PD patients after receiving LFS and VFS was significantly better than at the baseline and with HFS (all p < 0.0001). Conclusion: Our findings indicate that HFS is ineffective at improving the majority of axial symptoms in advanced PD. However, both the LFS and VFS programming paradigms exhibit significant improvements in various axial symptoms. [ABSTRACT FROM AUTHOR]