Saved in:
| Main Authors: | , , , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2026
|
| Subjects: | |
| Online Access: | https://arxiv.org/abs/2601.16011 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866917217784823808 |
|---|---|
| author | Forgaard, Theodor Reksten, Jarle H. Waldeland, Anders U. Marsocci, Valerio Longépé, Nicolas Kampffmeyer, Michael Salberg, Arnt-Børre |
| author_facet | Forgaard, Theodor Reksten, Jarle H. Waldeland, Anders U. Marsocci, Valerio Longépé, Nicolas Kampffmeyer, Michael Salberg, Arnt-Børre |
| contents | Current Earth observation foundation models are architecturally rigid, struggle with heterogeneous sensors and are constrained to fixed patch sizes. This limits their deployment in real-world scenarios requiring flexible computeaccuracy trade-offs. We propose THOR, a "computeadaptive" foundation model that solves both input heterogeneity and deployment rigidity. THOR is the first architecture to unify data from Copernicus Sentinel-1, -2, and -3 (OLCI & SLSTR) satellites, processing their native 10 m to 1000 m resolutions in a single model. We pre-train THOR with a novel randomized patch and input image size strategy. This allows a single set of pre-trained weights to be deployed at inference with any patch size, enabling a dynamic trade-off between computational cost and feature resolution without retraining. We pre-train THOR on THOR Pretrain, a new, large-scale multi-sensor dataset and demonstrate state-of-the-art performance on downstream benchmarks, particularly in data-limited regimes like the PANGAEA 10% split, validating that THOR's flexible feature generation excels for diverse climate and society applications. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2601_16011 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | THOR: A Versatile Foundation Model for Earth Observation Climate and Society Applications Forgaard, Theodor Reksten, Jarle H. Waldeland, Anders U. Marsocci, Valerio Longépé, Nicolas Kampffmeyer, Michael Salberg, Arnt-Børre Image and Video Processing Artificial Intelligence Current Earth observation foundation models are architecturally rigid, struggle with heterogeneous sensors and are constrained to fixed patch sizes. This limits their deployment in real-world scenarios requiring flexible computeaccuracy trade-offs. We propose THOR, a "computeadaptive" foundation model that solves both input heterogeneity and deployment rigidity. THOR is the first architecture to unify data from Copernicus Sentinel-1, -2, and -3 (OLCI & SLSTR) satellites, processing their native 10 m to 1000 m resolutions in a single model. We pre-train THOR with a novel randomized patch and input image size strategy. This allows a single set of pre-trained weights to be deployed at inference with any patch size, enabling a dynamic trade-off between computational cost and feature resolution without retraining. We pre-train THOR on THOR Pretrain, a new, large-scale multi-sensor dataset and demonstrate state-of-the-art performance on downstream benchmarks, particularly in data-limited regimes like the PANGAEA 10% split, validating that THOR's flexible feature generation excels for diverse climate and society applications. |
| title | THOR: A Versatile Foundation Model for Earth Observation Climate and Society Applications |
| topic | Image and Video Processing Artificial Intelligence |
| url | https://arxiv.org/abs/2601.16011 |