Deep Spatially-Regularized and Superpixel-Based Diffusion Learning for Unsupervised Hyperspectral Image Clustering

2026-04-14

AI Summary

This study addresses the challenge of effectively integrating spatial and spectral information in unsupervised hyperspectral image clustering, a limitation that often constrains clustering accuracy. To overcome this, the authors propose the DS²DL algorithm, which first employs an unsupervised masked autoencoder (UMAE) based on the Vision Transformer architecture to learn denoised latent representations. Subsequently, it constructs a spatially regularized diffusion graph guided by entropy rate superpixels (ERS) and performs clustering in the compressed latent space. This approach uniquely combines masked autoencoder pretraining with superpixel-constrained diffusion graphs to more accurately capture the intrinsic geometric structure of the data manifold. Experimental results demonstrate that DS²DL significantly improves clustering accuracy and overall performance on the Botswana and Kennedy Space Center (KSC) hyperspectral datasets.

Technology Category

• Machine Learning: Clustering
• Computer Vision: Representation Learning for Vision
• Constraint Satisfaction and Optimization: Distributed CSP/Optimization

Application Category

• Graph Algorithms and Modeling for the Web: Graph embeddings and representation learning for Web-related graphs
• Search and Retrieval-Augmented AI: Vertical and domain-specific search
• Web Mining and Content Analysis: Normalization, clustering, classification, and summarization of Web text

Abstract

An unsupervised framework for hyperspectral image (HSI) clustering is proposed that incorporates masked deep representation learning with diffusion-based clustering, extending the Spatially-Regularized Superpixel-based Diffusion Learning ($S^2DL$) algorithm. Initially, a denoised latent representation of the original HSI is learned via an unsupervised masked autoencoder (UMAE) model with a Vision Transformer backbone. The UMAE takes spatial context and long-range spectral correlations into account and incorporates an efficient pretraining process via masking that utilizes only a small subset of training pixels. In the next stage, the entropy rate superpixel (ERS) algorithm is used to segment the image into superpixels, and a spatially regularized diffusion graph is constructed using Euclidean and diffusion distances within the compressed latent space instead of the HSI space. The proposed algorithm, Deep Spatially-Regularized Superpixel-based Diffusion Learning ($DS^2DL$), leverages more faithful diffusion distances and subsequent diffusion graph construction that better reflect the intrinsic geometry of the underlying data manifold, improving labeling accuracy and clustering quality. Experiments on Botswana and KSC datasets demonstrate the efficacy of $DS^2DL$.

Problem

• Hyperspectral Image Clustering
• Unsupervised Learning
• Spatial Regularization
• Superpixel Segmentation
• Diffusion Learning

Innovation

• masked autoencoder
• diffusion learning
• superpixel segmentation
• spatial regularization
• hyperspectral image clustering