Multi-temporal Calving Front Segmentation

2025-12-12

AI Summary

Seasonal snow cover and supraglacial debris in SAR imagery cause instability in calving front segmentation for marine-terminating glaciers. To address this, we propose a multi-temporal collaborative modeling framework. Our approach innovatively introduces an inter-frame feature map temporal information exchange mechanism and is the first to embed multi-temporal modeling into the Tyrion architecture, augmented with a spatiotemporal feature alignment and interaction module. Evaluated on the CaFFe benchmark, our method achieves a mean distance error of 184.4 m and a mean intersection-over-union (IoU) of 83.6%, establishing new state-of-the-art performance. The proposed framework significantly enhances the robustness and generalizability of calving front localization under challenging SAR conditions. It provides a scalable technical foundation for operational remote sensing monitoring of glacier dynamics.

Technology Category

• Computer Vision: Remote Sensing / Geospatial AI
• Knowledge Representation and Reasoning: Geometric, Spatial, and Temporal Reasoning
• Machine Learning: Calibration & Uncertainty Quantification

Application Category

• Search and Retrieval-Augmented AI: Retrieval-Augmented Generation (RAG) and multi-modal RAG
• Graph Algorithms and Modeling for the Web: Efficient manipulation of static and dynamic Web-related graphs
• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs

Abstract

The calving fronts of marine-terminating glaciers undergo constant changes. These changes significantly affect the glacier's mass and dynamics, demanding continuous monitoring. To address this need, deep learning models were developed that can automatically delineate the calving front in Synthetic Aperture Radar imagery. However, these models often struggle to correctly classify areas affected by seasonal conditions such as ice melange or snow-covered surfaces. To address this issue, we propose to process multiple frames from a satellite image time series of the same glacier in parallel and exchange temporal information between the corresponding feature maps to stabilize each prediction. We integrate our approach into the current state-of-the-art architecture Tyrion and accomplish a new state-of-the-art performance on the CaFFe benchmark dataset. In particular, we achieve a Mean Distance Error of 184.4 m and a mean Intersection over Union of 83.6.

Problem

• Automatically segment glacier calving fronts in SAR imagery
• Improve accuracy in seasonal conditions like ice melange
• Enhance predictions using multi-temporal satellite image analysis

Innovation

• Parallel processing of multi-temporal satellite imagery
• Temporal information exchange between feature maps
• Integration into Tyrion architecture for enhanced accuracy