Deep Global Clustering for Hyperspectral Image Segmentation: Concepts, Applications, and Open Challenges
This addresses computational bottlenecks in hyperspectral imaging for domain-specific applications like agricultural monitoring, but it is incremental due to optimization instability.
The paper tackles hyperspectral image segmentation by proposing Deep Global Clustering (DGC), a memory-efficient framework that learns global clustering from local patches without pre-training, achieving a mean IoU of 0.925 for background-tissue separation on a leaf disease dataset.
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.