Expert Kernel Generation Network Driven by Contextual Mapping for Hyperspectral Image Classification
Guandong Li, Mengxia Ye
TL;DR
This work tackles hyperspectral image classification under challenges of high dimensionality and sparse ground objects by introducing EKGNet, a 3D-DenseNet–based architecture augmented with a context-aware mapping network and a dynamic kernel generation module. The context-aware mapping network produces input-specific attention weights $\alpha$ over $K$ base kernels via $\alpha = \text{Softmax}(f_{ ext{map}}(g)/\tau)$ with $g = \text{AvgPool3d}(X)$, enabling a dynamic kernel $W_{ ext{dyn}} = \sum_{k=1}^K \alpha_k W_k$ for each sample. The two components are tightly integrated: the mapping network guides kernel assembly, while the expert convolution system realizes adaptive, region-specific feature extraction, improving robustness to spectral redundancy and sparse targets. The model incorporates two design innovations for 3D-DenseNet—an exponentially increasing growth rate and fully dense connectivity—to boost feature reuse and efficiency without increasing depth or width. Evaluations on Indian Pines, Pavia University, and Kennedy Space Center show state-of-the-art accuracy, with faster convergence and competitive computational efficiency, indicating strong practical impact for hyperspectral land-cover classification.
Abstract
Deep neural networks face several challenges in hyperspectral image classification, including high-dimensional data, sparse distribution of ground objects, and spectral redundancy, which often lead to classification overfitting and limited generalization capability. To more efficiently adapt to ground object distributions while extracting image features without introducing excessive parameters and skipping redundant information, this paper proposes EKGNet based on an improved 3D-DenseNet model, consisting of a context-aware mapping network and a dynamic kernel generation module. The context-aware mapping module translates global contextual information of hyperspectral inputs into instructions for combining base convolutional kernels, while the dynamic kernels are composed of K groups of base convolutions, analogous to K different types of experts specializing in fundamental patterns across various dimensions. The mapping module and dynamic kernel generation mechanism form a tightly coupled system - the former generates meaningful combination weights based on inputs, while the latter constructs an adaptive expert convolution system using these weights. This dynamic approach enables the model to focus more flexibly on key spatial structures when processing different regions, rather than relying on the fixed receptive field of a single static convolutional kernel. EKGNet enhances model representation capability through a 3D dynamic expert convolution system without increasing network depth or width. The proposed method demonstrates superior performance on IN, UP, and KSC datasets, outperforming mainstream hyperspectral image classification approaches.