Recent advancements in geospatial AI are focusing on enhancing the accuracy and efficiency of earth observation and mapping techniques, addressing critical challenges in various applications. Notably, the development of billion-scale foundation models for semantic segmentation is improving the generalization of synthetic aperture radar imagery across diverse domains, which is essential for effective environmental monitoring and disaster response. Additionally, novel frameworks for phase unwrapping in interferometric synthetic aperture radar (InSAR) are enabling the analysis of complex deformation patterns, crucial for earthquake monitoring. The integration of advanced deep learning architectures for land cover mapping is facilitating better ecological assessments in riverine environments, while self-supervised learning approaches are streamlining the construction of high-definition maps for autonomous vehicles. Furthermore, innovative methods that leverage human mobility data are refining the understanding of points-of-interest, enhancing their representation for urban planning and navigation. Collectively, these developments are paving the way for more robust and scalable geospatial solutions across various sectors.
Synthetic Aperture Radar (SAR) enables global, all-weather earth observation. However, owing to diverse imaging mechanisms, domain shifts across sensors and regions severely hinder its semantic genera...
Phase unwrapping remains a critical and challenging problem in InSAR processing, particularly in scenarios involving complex deformation patterns. In earthquake-related deformation, shallow sources ca...
Accurate land cover mapping in riverine environments is essential for effective river management, ecological understanding, and geomorphic change monitoring. This study explores the use of Point Trans...
Accurate digital surface models (DSMs) are essential for many geospatial applications, including urban monitoring, environmental analyses, infrastructure management, and change detection. However, lar...
Autonomous vehicles rely on map information to understand the world around them. However, the creation and maintenance of offline high-definition (HD) maps remains costly. A more scalable alternative ...
Recent advances in generative networks have enabled new approaches to subsurface velocity model synthesis, offering a compelling alternative to traditional methods such as Full Waveform Inversion. How...
Natural language provides an intuitive way to express spatial intent in geospatial applications. While existing localization methods often rely on dense point cloud maps or high-resolution imagery, Op...
The Mariana Trench and the Qinghai-Tibet Plateau exhibit significant similarities in geological origins and microbial metabolic functions. Given that deep-sea biological sampling faces prohibitive cos...
We present Brainstacks, a modular architecture for continual multi-domain fine-tuning of large language models that packages domain expertise as frozen adapter stacks composing additively on a shared ...
In many real-world settings, such as environmental monitoring, disaster response, or public health, with costly and difficult data collection and dynamic environments, strategically sampling from unob...