Poly2Vec: Polymorphic Fourier-Based Encoding of Geospatial Objects for GeoAI Applications

2024-08-27

AI Summary

To address the lack of unified, relation-aware representations for heterogeneous geospatial objects (points, polylines, polygons) in GeoAI, this paper proposes a polymorphic Fourier encoding framework. First, diverse geometric objects are serialized into Fourier frequency-domain representations. Second, a learnable magnitude-phase fusion module is designed to adaptively model core spatial relations—including global position, topology, orientation, and distance. Third, a spatial-relation-aware contrastive training strategy is introduced. The method consistently outperforms specialized encoders across five downstream tasks. When integrated into state-of-the-art GeoAI pipelines, it significantly improves performance in population prediction and land-use inference. This work establishes the first unified spatial object encoder that is cross-geometric-type invariant, task-agnostic, and explicitly models spatial relations.

Technology Category

• Knowledge Representation and Reasoning: Geometric, Spatial, and Temporal Reasoning
• Computer Vision: Remote Sensing / Geospatial AI
• Reasoning under Uncertainty: Relational Probabilistic Models

Application Category

• Graph Algorithms and Modeling for the Web: Algorithms and analysis for heterogeneous, signed, attributed, multi-relational, temporal, higher-order, and annotated Web-related graphs
• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs
• Search and Retrieval-Augmented AI: Web query analysis, representation and understanding

Abstract

Encoding geospatial objects is fundamental for geospatial artificial intelligence (GeoAI) applications, which leverage machine learning (ML) models to analyze spatial information. Common approaches transform each object into known formats, like image and text, for compatibility with ML models. However, this process often discards crucial spatial information, such as the object's position relative to the entire space, reducing downstream task effectiveness. Alternative encoding methods that preserve some spatial properties are often devised for specific data objects (e.g., point encoders), making them unsuitable for tasks that involve different data types (i.e., points, polylines, and polygons). To this end, we propose Poly2Vec, a polymorphic Fourier-based encoding approach that unifies the representation of geospatial objects, while preserving the essential spatial properties. Poly2Vec incorporates a learned fusion module that adaptively integrates the magnitude and phase of the Fourier transform for different tasks and geometries. We evaluate Poly2Vec on five diverse tasks, organized into two categories. The first empirically demonstrates that Poly2Vec consistently outperforms object-specific baselines in preserving three key spatial relationships: topology, direction, and distance. The second shows that integrating Poly2Vec into a state-of-the-art GeoAI workflow improves the performance in two popular tasks: population prediction and land use inference.

Problem

• Encodes geospatial objects preserving spatial properties for GeoAI
• Unifies representation of points, polylines, and polygons
• Improves performance in population prediction and land use inference

Innovation

• Polymorphic Fourier-based encoding for geospatial objects
• Learned fusion module integrates Fourier magnitude and phase
• Unified representation preserves topology, direction, and distance