Panopticon: Advancing Any-Sensor Foundation Models for Earth Observation
Fuente:
arXiv
Guardado en:
| Autores principales: | , , , , , , , , |
|---|---|
| Formato: | Preprint |
| Publicado: |
2025
|
| Materias: | |
| Acceso en línea: | |
| Etiquetas: |
Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
|
| _version_ | 1866909714858639360 |
|---|---|
| author | Waldmann, Leonard Shah, Ando Wang, Yi Lehmann, Nils Stewart, Adam J. Xiong, Zhitong Zhu, Xiao Xiang Bauer, Stefan Chuang, John |
| author_facet | Waldmann, Leonard Shah, Ando Wang, Yi Lehmann, Nils Stewart, Adam J. Xiong, Zhitong Zhu, Xiao Xiang Bauer, Stefan Chuang, John |
| contents | Earth observation (EO) data features diverse sensing platforms with varying spectral bands, spatial resolutions, and sensing modalities. While most prior work has constrained inputs to fixed sensors, a new class of any-sensor foundation models able to process arbitrary sensors has recently emerged. Contributing to this line of work, we propose Panopticon, an any-sensor foundation model built on the DINOv2 framework. We extend DINOv2 by (1) treating images of the same geolocation across sensors as natural augmentations, (2) subsampling channels to diversify spectral input, and (3) adding a cross attention over channels as a flexible patch embedding mechanism. By encoding the wavelength and modes of optical and synthetic aperture radar sensors, respectively, Panopticon can effectively process any combination of arbitrary channels. In extensive evaluations, we achieve state-of-the-art performance on GEO-Bench, especially on the widely-used Sentinel-1 and Sentinel-2 sensors, while out-competing other any-sensor models, as well as domain adapted fixed-sensor models on unique sensor configurations. Panopticon enables immediate generalization to both existing and future satellite platforms, advancing sensor-agnostic EO. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2503_10845 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | Panopticon: Advancing Any-Sensor Foundation Models for Earth Observation Waldmann, Leonard Shah, Ando Wang, Yi Lehmann, Nils Stewart, Adam J. Xiong, Zhitong Zhu, Xiao Xiang Bauer, Stefan Chuang, John Machine Learning Earth observation (EO) data features diverse sensing platforms with varying spectral bands, spatial resolutions, and sensing modalities. While most prior work has constrained inputs to fixed sensors, a new class of any-sensor foundation models able to process arbitrary sensors has recently emerged. Contributing to this line of work, we propose Panopticon, an any-sensor foundation model built on the DINOv2 framework. We extend DINOv2 by (1) treating images of the same geolocation across sensors as natural augmentations, (2) subsampling channels to diversify spectral input, and (3) adding a cross attention over channels as a flexible patch embedding mechanism. By encoding the wavelength and modes of optical and synthetic aperture radar sensors, respectively, Panopticon can effectively process any combination of arbitrary channels. In extensive evaluations, we achieve state-of-the-art performance on GEO-Bench, especially on the widely-used Sentinel-1 and Sentinel-2 sensors, while out-competing other any-sensor models, as well as domain adapted fixed-sensor models on unique sensor configurations. Panopticon enables immediate generalization to both existing and future satellite platforms, advancing sensor-agnostic EO. |
| title | Panopticon: Advancing Any-Sensor Foundation Models for Earth Observation |
| topic | Machine Learning |
| url | https://arxiv.org/abs/2503.10845 |