Abstract:Hyperspectral image (HSI) classification is an important topic in remote sensing tasks, aiming to exploit the spectral and spatial information in HSI to assign category labels to each pixel. Recently, numerous deep learning (DL) methods, such as convolutional neural network (CNN) and transformer, have been introduced and achieved promising results. These DL-based models tend to use patch-based input patterns, where a patch is extracted around a central pixel and the neighboring pixels serve as auxiliary roles. However, compared to developing novel models, the characteristics of the patch-based HSI classification framework have not been adequately explored. Most existing patch-based DL methods tend to overlook the potential relationships between the central pixel and its neighboring pixels. In addition, the unique spectral characteristics of HSI (e.g., adjacent bands possess higher correlation) require special attention. Moreover, the design of feature extractors for HSIs appears to be constrained primarily to the spatial and spectral domains. To cope with the above problems, we propose an HSI classification specifically designed model (HSIC-SDM), which encompasses three branches—frequency feature learning, spatial feature learning, and spectral feature learning. The frequency feature learning branch employs a Fourier-based feature extractor to acquire representations in the frequency domain. The spatial feature learning branch emphasizes the significance of central pixel and edge information. The spectral feature learning branch concerns the high correlations in the adjacent spectral bands. Multidomain features derived from three branches provide diverse perspectives for the final classification. Extensive experiments on three datasets demonstrate the superiority of the proposed method over the state-of-the-art (SOTA) compared methods. The code of our work will be available publicly at https://github.com/Tikiten/HSIC-SDM.

Exploring Hyperspectral Histopathology Image Segmentation from a Deformable Perspective

Super-Resolution Based Patch-Free 3D Medical Image Segmentation with Self-Supervised Guidance

S^3R: Self-supervised Spectral Regression for Hyperspectral Histopathology Image Classification

(SR)-R-3: Self-supervised Spectral Regression for Hyperspectral Histopathology Image Classification

Hyperspectral Image Classification Based on Multiscale SpectralSpatial Deformable Network

Hyperspectral Image Reconstruction from RGB Input Through Highlighting Intrinsic Properties

Hyperspectral Image Restoration with Self-Supervised Learning: A Two-Stage Training Approach

Hyperspectral Image Instance Segmentation Using Spectral–Spatial Feature Pyramid Network

SpecTr: Spectral Transformer for Microscopic Hyperspectral Pathology Image Segmentation

Factor Space and Spectrum for Medical Hyperspectral Image Segmentation

Exploring the Applicability of Spectral Recovery in Semantic Segmentation of RGB Images

SegHSI: Semantic Segmentation of Hyperspectral Images With Limited Labeled Pixels

An Unsupervised Dehazing Network With Hybrid Prior Constraints for Hyperspectral Image

HyperspectraI Image Classification with Diverse Region Multiscale Feature Extraction and Spectral Imaging

DSSNet: A Simple Dilated Semantic Segmentation Network for Hyperspectral Imagery Classification

A hyperspectral image classification method based on feature enhancement and a hybrid deformable convolution network

Fortifying Centers and Edges: Multidomain Feature Learning Meets Hyperspectral Image Classification

A Spectral Spatial Attention Fusion with Deformable Convolutional Residual Network for Hyperspectral Image Classification.

Spectral Memory-Enhanced Network with Local Non-local and Low-rank Priors for Hyperspectral Compressive Imaging

Robust deep learning-based semantic organ segmentation in hyperspectral images