Advances in Machine Learning and Deep Learning for Hyperspectral Image Classification 

About the Special Session

Hyperspectral image classification is a rapidly evolving field with significant relevance to industrial, environmental, and security applications. Hyperspectral imaging captures high-dimensional data across hundreds of spectral bands, enabling detailed analysis of materials, chemicals, and environmental features. However, extracting meaningful information from hyperspectral data is highly challenging due to the high dimensionality, unbalanced class distributions, limited labeled samples, spectral variability, and complex interactions with substrates or backgrounds.

The objective of this special session is to bring together researchers and practitioners in machine learning, deep learning, and remote sensing to share novel computational approaches for hyperspectral image classification. The session will focus on recent advances that enhance classification accuracy, robustness, and generalization in real-world scenarios, where traditional methods often fail. Specifically, it aims to highlight innovative solutions that leverage AI to process large-scale hyperspectral datasets efficiently, overcome class imbalance, handle unlabeled or weakly labeled data, and extract meaningful spectral and spatial features.

Key challenges in hyperspectral image classification include:

·         High dimensionality and spectral redundancy, which can lead to the “curse of dimensionality.”

·         Limited annotated data and class imbalance, making supervised learning difficult.

·         Variability in environmental conditions and substrates that alter spectral signatures.

·         Noise, atmospheric interference, and sensor-specific artifacts in real-world data.

Promising approaches to address these challenges include:

·         Deep learning architectures (CNNs, RNNs, Transformers) for spatio-spectral feature extraction.

·         Semi-supervised, self-supervised, and few-shot learning methods for limited labeled data.

·         Transfer learning and domain adaptation to generalize across different sensors and conditions.

·         Multi-view clustering to jointly exploit complementary spatial, spectral, and other feature representations for unsupervised hyperspectral image classification.

·         Data augmentation and generative models to expand training datasets and improve robustness.

·         Hybrid physics-informed machine learning that combines domain knowledge with data-driven models.

The aim of this special session is to provide a platform for exchanging cutting-edge research, discussing open challenges, and fostering collaboration in the development of AI-driven hyperspectral image classification techniques. Contributions will be particularly welcomed that demonstrate practical applications in industry, environmental monitoring, agriculture, chemical detection, or security, emphasizing how AI and deep learning can empower real-world decision-making. The session also encourages interactive discussions on datasets, reproducibility, evaluation metrics, and integration of hyperspectral AI methods into industrial workflows.

Key Information

• Date: June 21 - 26, 2026

• Venue: MECC Maastricht, the Netherlands

• Paper Submission Deadline: January 31, 2026

Call for Papers

This Special Session invites papers on the following topics, including but not limited to: 

Remote Sensing Applications of Hyperspectral Imaging

• Exploring the use of hyperspectral imaging in various remote sensing applications
  
  

• Leveraging hyperspectral data for improved classification and analysis in remote sensing
  
  

Hyperspectral Imaging for Advanced Remote Sensing Techniques

• Integration of hyperspectral imaging with other remote sensing technologies (e.g., LiDAR, radar)
  
  

• Novel approaches to data fusion and analysis for hyperspectral image classification in remote sensing

Deep Learning Architectures for Hyperspectral Image Classification

• Convolutional Neural Networks (CNNs)
  
  

• Generative Adversarial Networks (GANs)
  
  

• Recurrent Neural Networks (RNNs) and LSTMs for temporal hyperspectral data
  
  

• Transformer-based models for hyperspectral data analysis
  
  

Transfer Learning for Hyperspectral Image Classification

• Pretrained models for hyperspectral applications
  
  

• Domain adaptation techniques
  
  

• Fine-tuning approaches for hyperspectral datasets
  
  

Unsupervised and Semi-Supervised Learning Methods

• Clustering-based approaches for hyperspectral classification
  
  

• Self-supervised learning techniques
  
  

• Semi-supervised learning strategies using limited labeled data
  
  

Data Augmentation and Synthetic Data Generation for Hyperspectral Data

• Data augmentation techniques for deep learning in hyperspectral imagery
  
  

• Use of synthetic data and simulation-based methods for training deep learning models
  
  

Multisource and Multimodal Data Fusion

• Combining hyperspectral images with LiDAR, multispectral, or radar data for enhanced classification
  
  

• Multimodal learning approaches
  
  

Explainability and Interpretability in Hyperspectral Image Classification

• Techniques for improving model transparency and understanding
  
  

• Explainable AI (XAI) for hyperspectral imagery
  
  

Advanced Feature Extraction Techniques

• Feature selection and dimensionality reduction techniques for hyperspectral data (e.g., PCA, LDA)
  
  

• Spectral-spatial feature extraction models
  
  

Adversarial Attacks and Robustness in Hyperspectral Image Classification

• Evaluating model robustness against adversarial examples in hyperspectral images
  
  

• Techniques for improving the security and resilience of deep learning models
  
  

Real-Time Hyperspectral Image Classification

• Techniques for fast, real-time processing of hyperspectral data
  
  

• Efficient deep learning models for edge devices
  
  

Applications of Hyperspectral Image Classification

• Environmental monitoring and land cover classification
  
  

• Agricultural and crop analysis using hyperspectral data
  
  

• Urban planning and infrastructure monitoring with hyperspectral imaging
  
  

• Military and defense applications (e.g., target detection, reconnaissance)
  
  

Benchmarking and Evaluation of Deep Learning Models

• Establishing standard evaluation metrics for hyperspectral classification models
  
  

• Datasets and benchmarks for comparing new techniques
  
  

Hybrid and Ensemble Models for Hyperspectral Classification

• Combining multiple learning techniques (e.g., deep learning + traditional machine learning)
  
  

• Hybrid models for improved classification accuracy and efficiency
  
  

Hyperparameter Tuning and Optimization Techniques

• Best practices for hyperparameter optimization in hyperspectral deep learning models
  
  

• Bayesian optimization, grid search, and random search for tuning model performance
  
  

Challenges and Future Directions in Hyperspectral Image Classification

• Addressing issues like label scarcity, class imbalance, and noise in hyperspectral data
  
  

• Future trends and the role of AI in advancing hyperspectral image classification

Primary Contact

Nian Ashlee Zhang, Ph.D.

Professor of Electrical and Computer Engineering

Department of Electrical and Computer Engineering

School of Engineering and Applied Sciences (SEAS)

University of the District of Columbia

4200 Connecticut Avenue, NW, Washington, D.C. 20008 USA

Office: +1 (202) 274-6615

Email: nzhang@udc.edu

Organizers

Nian Ashlee Zhang, Ph.D. 

Professor of Electrical and Computer Engineering

Department of Electrical and Computer Engineering 

School of Engineering and Applied Sciences (SEAS) 

University of the District of Columbia

4200 Connecticut Avenue, NW, Washington, D.C. 20008 USA

Office: +1 (202) 274-6615

Email: nzhang@udc.edu

Webpage: https://www.udc.edu/directory/profiles/seas/nian-zhang

Man-Fai Leung, Ph.D.

Senior Lecturer

School of Computing and Information Science

Faculty of Science and Engineering

Anglia Ruskin University

Cambridge, U.K.

Email: man-fai.leung@aru.ac.uk

Webpage: https://www.aru.ac.uk/people/man-fai-leung

https://scholar.google.co.uk/citations?hl=en&user=auIXUW8AAAAJ

Feng Xia, Ph.D.

Professor of Data Science & AI

School of Computing Technologies

RMIT University

Melbourne, VIC 3000, Australia

Email: f.xia@ieee.org

Webpage: https://www.xia.ai/

Donald C. Wunsch II, Ph.D., MBA

Mary Finley Missouri Distinguished Professor

Director of Kummer Institute Center for AI and Autonomous Systems

Missouri University of Science and Technology

Rolla, MO 65409 USA

Email: dwunsch@mst.edu

Webpage: https://sites.mst.edu/wunschkicaias/

https://caias.mst.edu/

© 2025-2026 WCCI 2026. All Rights Reserved.