Deep Global Clustering for Hyperspectral Image Segmentation: Concepts, Applications, and Open Challenges

TL;DR

This paper introduces Deep Global Clustering (DGC), a memory-efficient, unsupervised framework for hyperspectral image segmentation that operates on small patches, enabling fast training and effective domain-specific analysis without pre-training.

Contribution

The paper presents DGC, a novel framework for hyperspectral image segmentation that learns global clustering structures from local patches without pre-training, addressing computational and transferability challenges.

Findings

DGC achieves a mean IoU of 0.925 in background-tissue separation.

DGC enables unsupervised disease detection with navigable semantic granularity.

Training is completed in under 30 minutes on consumer hardware.

Abstract

Hyperspectral imaging (HSI) analysis faces computational bottlenecks due to massive data volumes that exceed available memory. While foundation models pre-trained on large remote sensing datasets show promise, their learned representations often fail to transfer to domain-specific applications like close-range agricultural monitoring where spectral signatures, spatial scales, and semantic targets differ fundamentally. This report presents Deep Global Clustering (DGC), a conceptual framework for memory-efficient HSI segmentation that learns global clustering structure from local patch observations without pre-training. DGC operates on small patches with overlapping regions to enforce consistency, enabling training in under 30 minutes on consumer hardware while maintaining constant memory usage. On a leaf disease dataset, DGC achieves background-tissue separation (mean IoU 0.925) and demonstrates unsupervised disease detection through navigable semantic granularity. However, the framework suffers from optimization instability rooted in multi-objective loss balancing: meaningful representations emerge rapidly but degrade due to cluster over-merging in feature space. We position this work as intellectual scaffolding - the design philosophy has merit, but stable implementation requires principled approaches to dynamic loss balancing. Code and data are available at https://github.com/b05611038/HSI_global_clustering.