HieraRS: A Hierarchical Segmentation Paradigm for Remote Sensing Enabling Multi-Granularity Interpretation and Cross-Domain Transfer

2025-07-11

AI Summary

Existing remote sensing land cover and land use (LCLU) classification methods suffer from two key limitations: (1) flat modeling strategies fail to generate multi-granularity predictions aligned with the inherent tree-like hierarchical structure of LCLU taxonomies; and (2) cross-domain transfer studies predominantly focus on homogeneous scenarios (e.g., sensor or geographic shifts), lacking adaptability to heterogeneous hierarchical tasks (e.g., LCLU → crop classification). To address these, we propose HieraRS—the first end-to-end hierarchical segmentation paradigm tailored for remote sensing imagery. It introduces a Bidirectional Hierarchical Consistency Constraint Mechanism (BHCCM) to enforce semantic consistency across granularities and a TransLU dual-branch framework to enable cross-domain knowledge sharing and semantic alignment. HieraRS is the first method supporting efficient, robust transfer of LCLU models to heterogeneous hierarchical tasks. Evaluated on our large-scale benchmark MM-5B, it significantly improves both classification accuracy and hierarchical consistency, while demonstrating strong generalization to dynamic class expansion, cross-sensor, cross-scene, and cross-task settings.

Technology Category

• Computer Vision: Remote Sensing / Geospatial AI
• Machine Learning: Transfer, Domain Adaptation, Multi-Task Learning
• Intelligent Robots: Multimodal Perception & Sensor Fusion

Application Category

• Search and Retrieval-Augmented AI: Vertical and domain-specific search
• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs
• Graph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs

Abstract

Hierarchical land cover and land use (LCLU) classification aims to assign pixel-wise labels with multiple levels of semantic granularity to remote sensing (RS) imagery. However, existing deep learning-based methods face two major challenges: 1) They predominantly adopt a flat classification paradigm, which limits their ability to generate end-to-end multi-granularity hierarchical predictions aligned with tree-structured hierarchies used in practice. 2) Most cross-domain studies focus on performance degradation caused by sensor or scene variations, with limited attention to transferring LCLU models to cross-domain tasks with heterogeneous hierarchies (e.g., LCLU to crop classification). These limitations hinder the flexibility and generalization of LCLU models in practical applications. To address these challenges, we propose HieraRS, a novel hierarchical interpretation paradigm that enables multi-granularity predictions and supports the efficient transfer of LCLU models to cross-domain tasks with heterogeneous tree-structured hierarchies. We introduce the Bidirectional Hierarchical Consistency Constraint Mechanism (BHCCM), which can be seamlessly integrated into mainstream flat classification models to generate hierarchical predictions, while improving both semantic consistency and classification accuracy. Furthermore, we present TransLU, a dual-branch cross-domain transfer framework comprising two key components: Cross-Domain Knowledge Sharing (CDKS) and Cross-Domain Semantic Alignment (CDSA). TransLU supports dynamic category expansion and facilitates the effective adaptation of LCLU models to heterogeneous hierarchies. In addition, we construct MM-5B, a large-scale multi-modal hierarchical land use dataset featuring pixel-wise annotations. The code and MM-5B dataset will be released at: https://github.com/AI-Tianlong/HieraRS.

Problem

• Enables multi-granularity hierarchical predictions for remote sensing imagery
• Addresses cross-domain transfer challenges with heterogeneous tree-structured hierarchies
• Improves semantic consistency and classification accuracy in land cover models

Innovation

• Hierarchical segmentation for multi-granularity remote sensing
• Bidirectional Hierarchical Consistency Constraint Mechanism
• Dual-branch cross-domain transfer framework TransLU