Dual-Branch Residual Network for Cross-Domain Few-Shot Hyperspectral Image Classification with Refined Prototype
This work addresses domain adaptation for hyperspectral image classification in few-shot scenarios, which is an incremental improvement over existing metric-based methods.
The paper tackles the problem of cross-domain few-shot hyperspectral image classification by addressing computational costs, limited generalization, and domain shifts, achieving superior performance on four datasets compared to other methods.
Convolutional neural networks (CNNs) are effective for hyperspectral image (HSI) classification, but their 3D convolutional structures introduce high computational costs and limited generalization in few-shot scenarios. Domain shifts caused by sensor differences and environmental variations further hinder cross-dataset adaptability. Metric-based few-shot learning (FSL) prototype networks mitigate this problem, yet their performance is sensitive to prototype quality, especially with limited samples. To overcome these challenges, a dual-branch residual network that integrates spatial and spectral features via parallel branches is proposed in this letter. Additionally, more robust refined prototypes are obtained through a regulation term. Furthermore, a kernel probability matching strategy aligns source and target domain features, alleviating domain shift. Experiments on four publicly available HSI datasets illustrate that the proposal achieves superior performance compared to other methods.