Pan-Arctic Permafrost Landform and Human-built Infrastructure Feature Detection with Vision Transformers and Location Embeddings

2025-06-03

AI Summary

Semantic segmentation of Arctic periglacial landforms—particularly ice-wedge polygons and thermokarst slumps—faces challenges including severe label scarcity, poor cross-regional generalization, and high spectral variability in large-scale remote sensing mapping of permafrost terrain and anthropogenic infrastructure. Method: We propose a remote sensing semantic segmentation framework integrating geospatial position embeddings with a Vision Transformer (ViT). Crucially, we inject latitude–longitude encodings directly into the ViT feature space to jointly model high-latitude spectral diversity and spatial structure; we further augment the approach with self-supervised pretraining and multimodal feature fusion to mitigate data scarcity. Results: Evaluated on pan-Arctic sub-meter satellite imagery, our method achieves F1-scores of 0.92 (+0.08) for ice-wedge polygon detection and 0.87 (+0.11) for thermokarst slump detection—substantially outperforming CNN-based baselines. This work provides the first empirical validation of spatially aware Transformers for robust and effective polar remote sensing interpretation.

Technology Category

• Computer Vision: Remote Sensing / Geospatial AI
• Intelligent Robots: Multimodal Perception & Sensor Fusion
• Knowledge Representation and Reasoning: Geometric, Spatial, and Temporal Reasoning

Application Category

• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs
• Search and Retrieval-Augmented AI: Vertical and domain-specific search
• Systems and Infrastructure for Web, Mobile and WoT: Applied ML and AI for Web-based mobile applications

Abstract

Accurate mapping of permafrost landforms, thaw disturbances, and human-built infrastructure at pan-Arctic scale using sub-meter satellite imagery is increasingly critical. Handling petabyte-scale image data requires high-performance computing and robust feature detection models. While convolutional neural network (CNN)-based deep learning approaches are widely used for remote sensing (RS),similar to the success in transformer based large language models, Vision Transformers (ViTs) offer advantages in capturing long-range dependencies and global context via attention mechanisms. ViTs support pretraining via self-supervised learning-addressing the common limitation of labeled data in Arctic feature detection and outperform CNNs on benchmark datasets. Arctic also poses challenges for model generalization, especially when features with the same semantic class exhibit diverse spectral characteristics. To address these issues for Arctic feature detection, we integrate geospatial location embeddings into ViTs to improve adaptation across regions. This work investigates: (1) the suitability of pre-trained ViTs as feature extractors for high-resolution Arctic remote sensing tasks, and (2) the benefit of combining image and location embeddings. Using previously published datasets for Arctic feature detection, we evaluate our models on three tasks-detecting ice-wedge polygons (IWP), retrogressive thaw slumps (RTS), and human-built infrastructure. We empirically explore multiple configurations to fuse image embeddings and location embeddings. Results show that ViTs with location embeddings outperform prior CNN-based models on two of the three tasks including F1 score increase from 0.84 to 0.92 for RTS detection, demonstrating the potential of transformer-based models with spatial awareness for Arctic RS applications.

Problem

• Detect permafrost landforms and human infrastructure using satellite imagery
• Improve model generalization for diverse Arctic spectral features
• Enhance feature detection with Vision Transformers and location embeddings

Innovation

• Vision Transformers for Arctic feature detection
• Location embeddings enhance model adaptation
• Self-supervised learning addresses data scarcity