Scalable Context-Preserving Model-Aware Deep Clustering for Hyperspectral Images

2025-06-12

AI Summary

Existing deep subspace clustering methods for unsupervised hyperspectral image (HSI) clustering suffer from high computational complexity (O(n²)), fragmented modeling of local spatial and non-local spectral structures, and insufficient end-to-end structural supervision. Method: We propose a scalable, one-stage, end-to-end deep clustering framework that jointly models local spatial continuity and non-local spectral similarity. Our approach introduces a novel co-optimization mechanism integrating spatial smoothing constraints with mini-cluster-level spectral consistency constraints. It achieves the first O(n)-complexity structure-aware subspace clustering, ensuring contextual preservation and full-length structural supervision. Key components include basis-representation subspace learning, spatial-filter alignment, mini-cluster-level prediction refinement, and joint structural regularization. Contribution/Results: The method significantly outperforms state-of-the-art approaches on multiple real-world HSI benchmarks, enables efficient large-scale clustering, and is publicly available as open-source code.

Technology Category

• Machine Learning: Clustering
• Constraint Satisfaction and Optimization: Distributed CSP/Optimization
• Computer Vision: Other Foundations of Computer Vision

Application Category

• Graph Algorithms and Modeling for the Web: Representation, reconstruction, and subgraph or motif discovery in Web-related graphs
• Web Mining and Content Analysis: Normalization, clustering, classification, and summarization of Web text
• Semantics and Knowledge: Scalable techniques for the creation, curation, publication, maintenance, and consumption of large, Web-based, structured, reusable, knowledge graphs and ontologies

Abstract

Subspace clustering has become widely adopted for the unsupervised analysis of hyperspectral images (HSIs). Recent model-aware deep subspace clustering methods often use a two-stage framework, involving the calculation of a self-representation matrix with complexity of O(n^2), followed by spectral clustering. However, these methods are computationally intensive, generally incorporating solely either local or non-local spatial structure constraints, and their structural constraints fall short of effectively supervising the entire clustering process. We propose a scalable, context-preserving deep clustering method based on basis representation, which jointly captures local and non-local structures for efficient HSI clustering. To preserve local structure (i.e., spatial continuity within subspaces), we introduce a spatial smoothness constraint that aligns clustering predictions with their spatially filtered versions. For non-local structure (i.e., spectral continuity), we employ a mini-cluster-based scheme that refines predictions at the group level, encouraging spectrally similar pixels to belong to the same subspace. Notably, these two constraints are jointly optimized to reinforce each other. Specifically, our model is designed as an one-stage approach in which the structural constraints are applied to the entire clustering process. The time and space complexity of our method is O(n), making it applicable to large-scale HSI data. Experiments on real-world datasets show that our method outperforms state-of-the-art techniques. Our code is available at: https://github.com/lxlscut/SCDSC

Problem

• Reduces high computational complexity in HSI clustering
• Integrates local and non-local spatial structure constraints
• Improves scalability for large-scale hyperspectral image data

Innovation

• Basis representation for scalable deep clustering
• Joint local and non-local structure optimization
• One-stage approach with O(n) complexity