HFC-SST: improved spatial-spectral transformer for hyperspectral few-shot classification

Huang, Zhiquan; Tang, Haojin; Li, Yanshan; Xie, Weixin

doi:10.1117/1.jrs.17.026509

Cited by 6 publications

(4 citation statements)

References 30 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Their method achieved state-of-the-art results on few-shot hyperspectral tasks using public datasets, demonstrating the potential of transformers to advance few-shot learning in this domain. Huang et al (2023) also recognized limitations of CNN-based models for few-shot hyperspectral image classification. They highlighted the inherent difficulty of CNNs in effectively capturing long-range spatial-spectral dependencies, especially in scenarios with limited training data.…”

Section: Few-shot Learning In Hyperspectral Images Classificationmentioning

confidence: 98%

Unlocking the capabilities of explainable few-shot learning in remote sensing

Lee,

Dam,

Ferdaus

et al. 2024

Artif Intell Rev

View full text Add to dashboard Cite

Recent advancements have significantly improved the efficiency and effectiveness of deep learning methods for image-based remote sensing tasks. However, the requirement for large amounts of labeled data can limit the applicability of deep neural networks to existing remote sensing datasets. To overcome this challenge, few-shot learning has emerged as a valuable approach for enabling learning with limited data. While previous research has evaluated the effectiveness of few-shot learning methods on satellite-based datasets, little attention has been paid to exploring the applications of these methods to datasets obtained from Unmanned Aerial Vehicles (UAVs), which are increasingly used in remote sensing studies. In this review, we provide an up-to-date overview of both existing and newly proposed few-shot classification techniques, along with appropriate datasets that are used for both satellite-based and UAV-based data. We demonstrate few-shot learning can effectively handle the diverse perspectives in remote sensing data. As an example application, we evaluate state-of-the-art approaches on a UAV disaster scene dataset, yielding promising results. Furthermore, we highlight the significance of incorporating explainable AI (XAI) techniques into few-shot models. In remote sensing, where decisions based on model predictions can have significant consequences, such as in natural disaster response or environmental monitoring, the transparency provided by XAI is crucial. Techniques like attention maps and prototype analysis can help clarify the decision-making processes of these complex models, enhancing their reliability. We identify key challenges including developing flexible few-shot methods to handle diverse remote sensing data effectively. This review aims to equip researchers with an improved understanding of few-shot learning’s capabilities and limitations in remote sensing, while pointing out open issues to guide progress in efficient, reliable and interpretable data-efficient techniques.

show abstract

Section: Few-shot Learning In Hyperspectral Images Classificationmentioning

confidence: 98%

Unlocking the capabilities of explainable few-shot learning in remote sensing

Lee,

Dam,

Ferdaus

et al. 2024

Artif Intell Rev

View full text Add to dashboard Cite

show abstract

“…The self-attention mechanism is a typical spatial attention module that has recently become well known in image processing. The self-attention mechanism captures long-range contextual information to obtain discriminative feature representations 54 . The self-attention mechanism works by establishing global dependencies through relationships between input sequences.…”

Section: Related Workmentioning

confidence: 99%

“…The self-attention mechanism captures long-range contextual information to obtain discriminative feature representations. 54 The self-attention mechanism works by establishing global dependencies through relationships between input sequences. Sun et al 28 and Li et al 29 used a self-attention mechanism to suppress the influence of interference pixels.…”

Section: Introductionmentioning

confidence: 99%

Center-similarity spectral-spatial attention network for hyperspectral image classification

Zhang,

Liang,

Niu

et al. 2024

J. Appl. Rem. Sens.

View full text Add to dashboard Cite

Hyperspectral image (HSI) classification aims to assign labels to pixels to be classified. The high-dimensional form of HSI and the introduction of spatial information can introduce challenges such as redundant spectral bands and interference pixels. Recently, many methods based on convolutional neural networks and attention mechanisms have been employed to address these issues. However, existing methods often fail to adequately utilize spatial information and exhibit interference pixel diffusion, which may result in these methods being unable to extract discriminative features. In addition, convolutions are not capable of extracting robust features, which results in the poor robustness of recent methods when HSI is rotated. To address these challenges, a center-similarity spectral-spatial attention network model (CS 3 AN) is proposed. First, a center-similarity spectral attention (CSSpeA) module is proposed to exploit the input spatial information rationally to better reduce the impact of redundant bands. Next, a center-similarity rectified spatial attention (CSRSpaA) module that selectively weighs neighboring pixels based on interference pixels to prevent the diffusion phenomenon is provided. Second, a similarity spatial aggregation module is employed to extract robust spectral-spatial features. Finally, the robust features are input into the softmax to get the label. The validity of the CS 3 AN is validated on three different databases.

show abstract

“…When the feature types in the classification scenario are very complex, the classification results often have the problem of spatial homogeneity and heterogeneity, for which many solutions have been proposed to incorporate spatial information in the classification [42]- [44]. Su et al [45] proposed a collaborative representational classification model with multifeature fusion dictionary learning (MCRC-DL), which simultaneously considers the spectral, local, global, and morphological features of the data and morphological features, which in turn yields the representation coefficients to determine the predicted class.…”

Section: Introductionmentioning

confidence: 99%

Mean-Weighted Collaborative Representation-Based Spatial-Spectral Joint Classification for Hyperspectral Images

Su,

Shi,

Xue

et al. 2024

IEEE J. Sel. Top. Appl. Earth Observations Remote Sensing

View full text Add to dashboard Cite

Collaborative representation (CR) models have been widely used in hyperspectral image (HSI) classification tasks. However, most collaborative representation classification models lack stability and generalization when targeting small samples as well as spatial homogeneity and heterogeneity problems. Therefore, this paper proposes a mean-weighted collaborative representation classification model (MWCRC) based on the joint spatial-spectral data. It imposes mean and weighted constraints on the representation coefficients based on CR, which attenuates the noise effect and increases the distinguishability between classes. Second, a sample augmentation method based on the principle of minimizing the representation residuals is proposed. Sample augmentation is realized through initial classification and calculation of representation residuals to achieve the objective of consolidating model stability and improving classification accuracy. Meanwhile, in order to alleviate the problem of spatial homogeneity and heterogeneity, the extended morphological profile (EMP) and the stacking approach are utilized to construct the joint spatial-spectral data for the classification of MWCRC. The superiority of the proposed method is demonstrated by experimental validation using a small number of training samples in three real datasets.

show abstract

HFC-SST: improved spatial-spectral transformer for hyperspectral few-shot classification

Cited by 6 publications

References 30 publications

Unlocking the capabilities of explainable few-shot learning in remote sensing

Unlocking the capabilities of explainable few-shot learning in remote sensing

Center-similarity spectral-spatial attention network for hyperspectral image classification

Mean-Weighted Collaborative Representation-Based Spatial-Spectral Joint Classification for Hyperspectral Images

Contact Info

Product

Resources

About