Curriculum Multi-Task Self-Supervision Improves Lightweight Architectures for Onboard Satellite Hyperspectral Image Segmentation

2025-09-16

AI Summary

To address the substantial data transmission burden and urgent demand for model lightweighting in onboard hyperspectral image (HSI) segmentation, this paper proposes a curriculum-based multi-task self-supervised learning framework. The method innovatively integrates masked image modeling with disentangled spatial-spectral jigsaw tasks, progressively increasing data complexity via a curriculum learning strategy to jointly capture spectral continuity, spatial structure, and global semantics within a unified framework. A lightweight network architecture enables efficient pretraining tailored to resource-constrained onboard environments. Extensive evaluation on four public HSI datasets demonstrates consistent improvements in downstream segmentation performance. Notably, the proposed model reduces parameter count by over 16,000× compared to state-of-the-art methods, significantly enhancing representational capacity and generalization of lightweight models for HSI analysis.

Technology Category

• Computer Vision: Segmentation
• Machine Learning: Multi-instance/Multi-view Learning
• Search and Optimization: Learning to Search

Application Category

• Web Mining and Content Analysis: Large pretrained models with web data
• Graph Algorithms and Modeling for the Web: Graph embeddings and representation learning for Web-related graphs
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking

Abstract

Hyperspectral imaging (HSI) captures detailed spectral signatures across hundreds of contiguous bands per pixel, being indispensable for remote sensing applications such as land-cover classification, change detection, and environmental monitoring. Due to the high dimensionality of HSI data and the slow rate of data transfer in satellite-based systems, compact and efficient models are required to support onboard processing and minimize the transmission of redundant or low-value data, e.g. cloud-covered areas. To this end, we introduce a novel curriculum multi-task self-supervised learning (CMTSSL) framework designed for lightweight architectures for HSI analysis. CMTSSL integrates masked image modeling with decoupled spatial and spectral jigsaw puzzle solving, guided by a curriculum learning strategy that progressively increases data complexity during self-supervision. This enables the encoder to jointly capture fine-grained spectral continuity, spatial structure, and global semantic features. Unlike prior dual-task SSL methods, CMTSSL simultaneously addresses spatial and spectral reasoning within a unified and computationally efficient design, being particularly suitable for training lightweight models for onboard satellite deployment. We validate our approach on four public benchmark datasets, demonstrating consistent gains in downstream segmentation tasks, using architectures that are over 16,000x lighter than some state-of-the-art models. These results highlight the potential of CMTSSL in generalizable representation learning with lightweight architectures for real-world HSI applications. Our code is publicly available at https://github.com/hugocarlesso/CMTSSL.

Problem

• Improving lightweight models for onboard satellite hyperspectral image segmentation
• Addressing high dimensionality and slow data transfer in satellite HSI systems
• Enabling efficient spatial and spectral reasoning with self-supervised learning

Innovation

• Curriculum multi-task self-supervised learning framework
• Integrates masked image modeling with jigsaw puzzles
• Trains lightweight encoders for satellite deployment