PulseExploreJournal ClubDebatesTrendingResearchersJournals
Instagram
HomeExploreJournal ClubTrending
Synapse
⌘+K
Synapse
February 8, 2026Remote Sensing0 citationsOpen Access

Contrastive–Transfer-Synergized Dual-Stream Transformer for Hyperspectral Anomaly Detection

LDLei DengJYJiaju YingQWQianghui Wang

Key Points

  • This research aims to improve hyperspectral anomaly detection by addressing feature representation and training inefficiencies.
  • Proposed a dual-stream transformer architecture integrating contrastive and transfer learning.
  • Utilized a spatial stream with adaptive elastic weight consolidation to prevent forgetting.
  • Employed a spectral stream with a variational autoencoder for data augmentation and feature extraction.
  • Fused spatial and spectral features for anomaly detection at the pixel level.
  • Implemented focal loss to manage class imbalance during training.
  • CTDST-HAD outperformed state-of-the-art methods in detection accuracy across various datasets.
  • Achieved significant improvements in efficiency, especially in challenging background conditions.
  • Maintained competitive inference speed while enhancing detection capabilities.

Abstract

Hyperspectral anomaly detection (HAD) aims to identify pixels that significantly differ from the background without prior knowledge. While deep learning-based reconstruction methods have shown promise, they often suffer from limited feature representation, inefficient training cycles, and sensitivity to imbalanced data distributions. To address these challenges, this paper proposes a novel contrastive–transfer-synergized dual-stream transformer for hyperspectral anomaly detection (CTDST-HAD). The framework integrates contrastive learning and transfer learning within a dual-stream architecture, comprising a spatial stream and a spectral stream, which are pre-trained separately and synergistically fine-tuned. Specifically, the spatial stream leverages general visual and hyperspectral-view datasets with adaptive elastic weight consolidation (EWC) to mitigate catastrophic forgetting. The spectral stream employs a variational autoencoder (VAE) enhanced with the RossThick–LiSparseR (R-L) physical-kernel-driven model for spectrally realistic data augmentation. During fine-tuning, spatial and spectral features are fused for pixel-level anomaly detection, with focal loss addressing class imbalance. Extensive experiments on nine real hyperspectral datasets demonstrate that CTDST-HAD outperforms state-of-the-art methods in detection accuracy and efficiency, particularly in complex backgrounds, while maintaining competitive inference speed.

Ask AI
Helpful
Bookmark
Share
View Full Paper

Cite This Study

Deng et al. (2026) studied this question.

synapsesocial.com/papers/6988278b0fc35cd7a88465e5https://doi.org/10.3390/rs18030516
Ask AI
Helpful
Bookmark
Share
View Full Paper