Earth Embeddings as Products: Taxonomy, Ecosystem, and Standardized Access
Fuente:
arXiv
Salvato in:
| Autori principali: | , , , , |
|---|---|
| Natura: | Preprint |
| Pubblicazione: |
2026
|
| Soggetti: | |
| Accesso online: | |
| Tags: |
Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
|
| _version_ | 1866914346063364096 |
|---|---|
| author | Fang, Heng Stewart, Adam J. Corley, Isaac Zhu, Xiao Xiang Azizpour, Hossein |
| author_facet | Fang, Heng Stewart, Adam J. Corley, Isaac Zhu, Xiao Xiang Azizpour, Hossein |
| contents | Geospatial Foundation Models (GFMs) provide powerful representations, but high compute costs hinder their widespread use. Pre-computed embedding data products offer a practical "frozen" alternative, yet they currently exist in a fragmented ecosystem of incompatible formats and resolutions. This lack of standardization creates an engineering bottleneck that prevents meaningful model comparison and reproducibility. We formalize this landscape through a three-layer taxonomy: Data, Tools, and Value. We survey existing products to identify interoperability barriers. To bridge this gap, we extend TorchGeo with a unified API that standardizes the loading and querying of diverse embedding products. By treating embeddings as first-class geospatial datasets, we decouple downstream analysis from model-specific engineering, providing a roadmap for more transparent and accessible Earth observation workflows. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2601_13134 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | Earth Embeddings as Products: Taxonomy, Ecosystem, and Standardized Access Fang, Heng Stewart, Adam J. Corley, Isaac Zhu, Xiao Xiang Azizpour, Hossein Software Engineering Computer Vision and Pattern Recognition Geospatial Foundation Models (GFMs) provide powerful representations, but high compute costs hinder their widespread use. Pre-computed embedding data products offer a practical "frozen" alternative, yet they currently exist in a fragmented ecosystem of incompatible formats and resolutions. This lack of standardization creates an engineering bottleneck that prevents meaningful model comparison and reproducibility. We formalize this landscape through a three-layer taxonomy: Data, Tools, and Value. We survey existing products to identify interoperability barriers. To bridge this gap, we extend TorchGeo with a unified API that standardizes the loading and querying of diverse embedding products. By treating embeddings as first-class geospatial datasets, we decouple downstream analysis from model-specific engineering, providing a roadmap for more transparent and accessible Earth observation workflows. |
| title | Earth Embeddings as Products: Taxonomy, Ecosystem, and Standardized Access |
| topic | Software Engineering Computer Vision and Pattern Recognition |
| url | https://arxiv.org/abs/2601.13134 |