A novel knowledge transfer paradigm from classification tasks to anomaly detection is proposed, which employs a pretrained convolutional enhanced Transformer (CET) classification network as the teacher model, transferring its rich spectral–spatial context encodings via feature-level distillation.
Abstract
Hyperspectral anomaly detection (HAD) aims to identify Earth’s surface anomalies without prior knowledge. However, most existing unsupervised HAD methods rely on training data roughly partitioned into background and anomalies, suffering from severe class imbalance and yielding feature representations heavily tied to scene-specific distributions. In contrast, supervised hyperspectral classification networks can extract well-organized spectral–spatial representations from diverse land-cover types; yet, this valuable knowledge remains underutilized in HAD due to the absence of labeled anomaly samples. To circumvent this limitation, this paper proposes a novel knowledge transfer paradigm from classification tasks to anomaly detection. The proposed framework employs a pretrained convolutional enhanced Transformer (CET) classification network as the teacher model, transferring its rich spectral–spatial context encodings via feature-level distillation. Concurrently, by synergizing latent feature alignment, Gaussian mixture model (GMM) priors, and reconstruction objectives, the student network is equipped with compact background modeling and highly discriminative representation abilities, which provide robust support for feature-space anomaly detection. Extensive experiments on spaceborne and airborne hyperspectral datasets demonstrate that our method substantially improves detection accuracy against complex terrestrial backgrounds with fast inference, while exhibiting superior cross-scene generalization.
Hyperspectral anomaly detection aims to identify pixels that significantly deviate from the surrounding background without prior knowledge of anomaly classes. Existing reconstruction-based methods usually learn background representations through autoencoders and detect anomalies using reconstruction residuals. However,...
A hyperspectral conditional diffusion for dense (HCD2) classification framework is proposed in this article, which reformulates HSI classification as a conditional label-map denoising process guided by hyperspectral features, enabling generative dense semantic inference.
Hong-Qiao Wang, Hai-Yan Sun, Xuan-Chen Yan et al.· IEEE Transactions on Geoscie...· 0 citations
A partial convolution autoencoder that employs 1D convolution layers to capture local spectral features to handle the complex data distribution of normal samples, and employs the nearest-neighbor similarity score computed against normal features stored in the memory bank as the anomaly score.
W. N. Khotimah, Moch Nafkhan Alzamzami, Malfino Muhammad Willianz et al.· IEEE Access· 0 citations
A self-adaptive background-attentive network with anomaly suppression (BASNet) is proposed, and a pyramid similarity downsampling module (PSDM) is employed to reduce the proportion of anomalies in the input image and improve multiscale background representation.
Xu Liu, Hao Yuan, Bo-Hui Tang et al.· IEEE Transactions on Neural...· 0 citations
This work proposes a four–stage hybrid framework for hyperspectral image classification that integrates spectral modeling and spatial feature extraction within a unified pipeline and compares the model with recent baselines, including DaViT.