adaptive band selection

This work addresses the limitations of existing hyperspectral band selection methods, which are sensitive to initialization, require a preset number of bands lacking flexibility, and exhibit unstable performance under spatially disjoint evaluation protocols. To overcome these issues, the authors propose SGBR-HC, a two-stage approach that first performs supervised band ranking based on class separability and spectral diversity to provide an informative prior for learnable sparse gating. In the second stage, the sparse gating module and a spatial classifier are jointly trained to adaptively determine the optimal number of bands. By integrating differentiable sparse gating, Hard-Concrete initialization, and spatially disjoint evaluation, the method effectively prevents information leakage. Experiments on the Pavia University and Houston 2013 datasets demonstrate state-of-the-art overall accuracy and Cohen’s kappa using only approximately 20 selected bands, while ablation studies confirm the critical role of the ranking prior.

Hyperspectral image (HSI) semantic segmentation suffers from spectral redundancy and high computational cost, while existing band selection methods often rely on preprocessing and are decoupled from downstream tasks. To address this, we propose Embedded Hyperspectral Band Selection (EHBS), the first end-to-end trainable framework that integrates band selection directly into semantic segmentation training. EHBS introduces a differentiable stochastic band gating mechanism for dynamic spectral filtering, employs a differentiable ℓ₀-norm regularization to precisely control band sparsity, and incorporates a parameter-free dynamic optimizer (DoG) that adaptively adjusts learning rates. Evaluated on two mainstream HSI segmentation benchmarks, EHBS achieves state-of-the-art performance—delivering higher accuracy and significantly reduced model complexity—while demonstrating strong generalization to grouped feature selection tasks.

Hyperspectral image (HSI) band redundancy incurs substantial computational overhead and information redundancy, impeding classification efficiency and accuracy. To address this, we propose a threshold-driven average correlation-based band selection method: a sliding-window average correlation model is constructed using Pearson correlation coefficients to quantify global inter-band dependencies; statistically significant thresholds are then employed to automatically identify an optimal subset of low-correlation, highly complementary bands—bypassing the complexity of clustering or iterative optimization. The method ensures both statistical robustness and discriminative power. Evaluated on the Pavia University and Salinas Valley datasets, it achieves classification accuracy comparable to state-of-the-art approaches while reducing input dimensionality by 40%–60% and accelerating inference speed by 2.1–3.4×, significantly enhancing model lightweighting and deployment efficiency.

Hyperspectral imaging (HSI) suffers from high spectral dimensionality and substantial redundancy, necessitating dimensionality reduction methods that jointly optimize predictive performance and interpretability. This paper proposes a post-hoc interpretable band selection framework for HSI: it introduces the deletion-insertion evaluation paradigm to HSI band selection for the first time, quantifying each band’s influence score on a pre-trained classifier’s decision. High-impact bands with clear physical meaning are then selected. The method integrates model-aligned contribution analysis, influence score aggregation, and empirical validation. Experiments on the Pavia University and Salinas datasets demonstrate that classifiers using only 30 selected bands achieve accuracy comparable to or exceeding that of full-spectrum models, while significantly reducing computational cost. Crucially, the approach ensures both interpretability—through transparent, physics-grounded band selection—and generalizability across diverse scenes and classifiers.

High-dimensional hyperspectral imaging (HSI) data suffer from severe band redundancy and high inter-band correlation, leading to the curse of dimensionality and degraded reconstruction accuracy. To address this, we propose a diversity-driven band grouping selection method based on Determinantal Point Processes (DPP), marking the first application of DPP to Earth observation (EO) band selection. We further integrate Spectral Angle Mapper (SAM) to quantify spectral similarity among bands, effectively mitigating group overlap and enabling physically interpretable, discriminative feature subset construction. Experiments demonstrate that our method achieves substantial dimensionality reduction—averaging over 60% band compression—while preserving critical spectral discriminability. Consequently, image reconstruction PSNR improves by 2.1–3.8 dB, and downstream classification and anomaly detection accuracy increases by 3.2–5.7%. This work establishes a novel paradigm for lightweight, high-accuracy intelligent remote sensing analysis.

This study addresses the sampling sensitivity of hyperspectral band selection and its unclear relationship with downstream segmentation performance. We evaluate the stability of six methods using Jaccard similarity and conduct Top-K band selection combined with algorithms such as Sim-LP on the Hyperspectral City dataset, benchmarked across three semantic segmentation models. Our analysis reveals that intra-method stability is informative yet not a reliable performance indicator, underscoring the necessity of comprehensive evaluation across samples, subsets, and models. Ultimately, the proposed approach achieves up to a 1.72 mF1 improvement alongside an 18–22× inference acceleration. These results demonstrate that efficient and competitive hyperspectral perception is attainable even with a limited number of spectral bands.

本文解决了高光谱图像分类中的维数约简与选择合适监督分类技术的问题，通过比较PCA、LDA及KNN、SVM、RF方法，发现PCA结合RF效果最佳。

This study addresses the critical challenge of wavelength selection in near-infrared spectroscopy for accurate and interpretable sugar content prediction, a task often hindered by noise sensitivity and instability in existing methods. To overcome these limitations, the authors propose the first application of combinatorial Bayesian optimization to this domain, formulating wavelength region selection as a binary black-box optimization problem. They construct a sparse quadratic surrogate model and solve the resulting quadratic unconstrained binary optimization (QUBO) problem using Thompson sampling combined with classical or quantum annealing. Compared to genetic algorithms and conventional simulated annealing, the proposed approach achieves significantly higher prediction accuracy. Moreover, the selected wavelength bands exhibit minimal root-mean-square error variation under single-bit perturbations, demonstrating superior consistency, robustness, and generalization capability.

该研究通过将稀疏矩阵视为二维信号并进行频域分析，探索了稀疏矩阵计算与谱分析之间的联系，以提高稀疏计算性能。

This study addresses the challenge of accurately classifying fish freshness from hyperspectral images, which is hindered by spectral dominance, ordinal label structure, and severe sample scarcity that limit the effectiveness of conventional deep learning approaches. To overcome these limitations, this work proposes SGNet, a lightweight network that introduces domain-aware spectral grouping convolution to decouple spectral and spatial features. By integrating depthwise separable convolution with a dual channel–spatial attention mechanism, SGNet enhances discriminative capability with minimal computational overhead. Evaluated on a self-collected dataset of salmon samples stored over 16 days, SGNet achieves a classification accuracy of 97.8% and an average absolute error of 0.64 days using only 4.75 million parameters—reducing parameter count by 5–18× compared to ResNet-50 and Vision Transformer while maintaining high accuracy and real-time performance.