Bridging the Style Gap: Style-Guided Distillation Domain Adaptation for Hyperspectral Image Classification

Land cover in different hyperspectral images (HSIs) commonly exhibits style differences and similarities in the same category and distinct categories. However, most of the existing cross-scene HSI classification methods overlook this issue and only conduct the feature-level alignment with unsupervised domain adaptation (UDA). To address this limitation, we propose the style-guided distillation domain adaptation (SGDDA) for HSI classification. First, a Fourier transform-based style transfer (FTST) module is proposed to generate an enhanced source HSI. It transfers stylistic features from the target domain (TD) to the source domain (SD) by substituting low-frequency components, thereby preserving semantic invariance while bridging their style gap. Second, the dual-path knowledge distillation (DPKD) module is designed to reduce ambiguity in category assignment for the SD. This is achieved through cross-domain consistency learning between original SD samples and their enhanced style-transferred counterparts, ensuring robust feature alignment across domains. Third, unlike existing methods that primarily utilize a classification threshold to select pseudo-label for target samples, we propose the confidence-aware dual pseudo-label consensus (CADPLC) strategy. This strategy dynamically selects the reliable pseudo-labels by leveraging both class prototype matching and teacher–student prediction consensus, eliminating reliance on fixed thresholds and significantly improving adaptation to the TD. Experiments on three benchmark datasets demonstrate the superiority of the proposed SGDDA in comparison with several state-of-the-art methods. The code is available at https://github.com/Wei-spvl/SGDDA

• Publication year

  2025

• Venue

  IEEE Transactions on Geoscience and Remote Sensing

• Publication date

  Unknown publication date

• Fields of study

  Computer Science, Engineering, Environmental Science

• Identifiers
  DOI 10.1109/TGRS.2025.3611972
• External record

  Open on Semantic Scholar

• Source metadata

  Semantic Scholar

• No claims are published for this paper.

• No concepts are published for this paper.

• Multilevel Prototype Alignment for Cross-Domain Few-Shot Hyperspectral Image Classification

  2025influential reference
• From Dynamic to Static: Stepwisely Generate HDR Image for Ghost Removal

  2025cited by this paper
• Cross-Scene Hyperspectral Image Classification With Consistency-Aware Customized Learning

  2025cited by this paper
• Dynamic semantic-geometric guidance and structure transfer network for cross-scene hyperspectral image classification

  2025cited by this paper
• Efficient Image Enhancement With a Diffusion-Based Frequency Prior

  2025cited by this paper
• Text-Driven Adaptive Semantic Alignment Network for Cross-Scene Hyperspectral Image Classification

  2025cited by this paper
• Adaptive Graph Modeling With Self-Training for Heterogeneous Cross-Scene Hyperspectral Image Classification

  2024cited by this paper
• StyleStudio: Text-Driven Style Transfer with Selective Control of Style Elements

  2024cited by this paper
• Classwise Prototype-Guided Alignment Network for Cross-Scene Hyperspectral Image Classification

  2024cited by this paper
• Masked Self-Distillation Domain Adaptation for Hyperspectral Image Classification

  2024cited by this paper
• Uncertainty estimation in HDR imaging with Bayesian neural networks

  2024cited by this paper
• UnmixDiff: Unmixing-Based Diffusion Model for Hyperspectral Image Synthesis

  2024cited by this paper
• Joint Adversarial Network With Semantic and Topology Fusion for Cross-Scene Hyperspectral Image Classification

  2024cited by this paper
• Cross-Domain Few-Shot Hyperspectral Image Classification With Cross-Modal Alignment and Supervised Contrastive Learning

  2024cited by this paper
• Spectral–Spatial Adversarial Multidomain Synthesis Network for Cross-Scene Hyperspectral Image Classification

  2024cited by this paper
• Adversarial Domain Adaptation Network With Calibrated Prototype and Dynamic Instance Convolution for Hyperspectral Image Classification

  2024cited by this paper
• Easy-to-Hard Domain Adaptation With Human Interaction for Hyperspectral Image Classification

  2024cited by this paper
• ALOFT: A Lightweight MLP-Like Architecture with Dynamic Low-Frequency Transform for Domain Generalization

  2023cited by this paper
• Supervised Contrastive Learning-Based Unsupervised Domain Adaptation for Hyperspectral Image Classification

  2023cited by this paper
• Dual-View Spectral and Global Spatial Feature Fusion Network for Hyperspectral Image Classification

  2023cited by this paper
• Prompt-based Distribution Alignment for Unsupervised Domain Adaptation

  2023cited by this paper
• Focus on Your Target: A Dual Teacher-Student Framework for Domain-adaptive Semantic Segmentation

  2023cited by this paper
• DiGA: Distil to Generalize and then Adapt for Domain Adaptive Semantic Segmentation

  2023cited by this paper
• DiffStyler: Controllable Dual Diffusion for Text-Driven Image Stylization

  2022cited by this paper
• Unsupervised Domain-Invariant Feature Learning for Cloud Detection of Remote Sensing Images

  2022cited by this paper
• Spatial-Temporal Based Multihead Self-Attention for Remote Sensing Image Change Detection

  2022cited by this paper
• Local Semantic Feature Aggregation-Based Transformer for Hyperspectral Image Classification

  2022cited by this paper
• Grouped Bidirectional LSTM Network and Multistage Fusion Convolutional Transformer for Hyperspectral Image Classification

  2022cited by this paper
• Two-Branch Attention Adversarial Domain Adaptation Network for Hyperspectral Image Classification

  2022cited by this paper
• Pivot-based Unsupervised Domain Adaptation for Pre-trained Language Model

  2022cited by this paper
• Estimating the Forage Neutral Detergent Fiber Content of Alpine Grassland in the Tibetan Plateau Using Hyperspectral Data and Machine Learning Algorithms

  2021influential reference
• Class-Wise Distribution Adaptation for Unsupervised Classification of Hyperspectral Remote Sensing Images

  2021cited by this paper
• Emerging Properties in Self-Supervised Vision Transformers

  2021cited by this paper
• DACS: Domain Adaptation via Cross-domain Mixed Sampling

  2020cited by this paper
• Content-Consistent Matching for Domain Adaptive Semantic Segmentation

  2020cited by this paper
• Deep HDR Imaging via A Non-Local Network

  2020cited by this paper
• Feature selection for cross-scene hyperspectral image classification using cross-domain ReliefF

  2019influential reference
• Momentum Contrast for Unsupervised Visual Representation Learning

  2019cited by this paper
• Remote sensing monitoring of multi-scale watersheds impermeability for urban hydrological evaluation

  2019influential reference
• 2018 IEEE GRSS Data Fusion Contest: Multimodal Land Use Classification [Technical Committees]

  2018cited by this paper
• A survey of landmine detection using hyperspectral imaging

  2017influential reference
• Squeeze-and-Excitation Networks

  2017cited by this paper
• Deep CORAL: Correlation Alignment for Deep Domain Adaptation

  2016cited by this paper
• Learning Transferable Features with Deep Adaptation Networks

  2015cited by this paper
• Hyperspectral and LiDAR Data Fusion: Outcome of the 2013 GRSS Data Fusion Contest

  2014cited by this paper
• Accelerating t-SNE using tree-based algorithms

  2014cited by this paper
• Multi-metric learning for multi-sensor fusion based classification

  2013cited by this paper
• Crop Yield Estimation Based on Unsupervised Linear Unmixing of Multidate Hyperspectral Imagery

  2013influential reference
• A Kernel Method for the Two-Sample-Problem

  2006cited by this paper
• FFTW: an adaptive software architecture for the FFT

  1998cited by this paper

• Multilevel Alignment-Based Domain Generalization Network for Cross-Scene Hyperspectral Image Classification

  2026cites this paper