Sat-JEPA-Diff: Bridging Self-Supervised Learning and Generative Diffusion for Remote Sensing
Fuente:
arXiv
Saved in:
| Main Authors: | , |
|---|---|
| Format: | Preprint |
| Published: |
2026
|
| Subjects: | |
| Online Access: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866914393420201984 |
|---|---|
| author | Komurcu, Kursat Petkevicius, Linas |
| author_facet | Komurcu, Kursat Petkevicius, Linas |
| contents | Predicting satellite imagery requires a balance between structural accuracy and textural detail. Standard deterministic methods like PredRNN or SimVP minimize pixel-based errors but suffer from the "regression to the mean" problem, producing blurry outputs that obscure subtle geographic-spatial features. Generative models provide realistic textures but often misleadingly reveal structural anomalies. To bridge this gap, we introduce Sat-JEPA-Diff, which combines Self-Supervised Learning (SSL) with Hidden Diffusion Models (LDM). An IJEPA module predicts stable semantic representations, which then route a frozen Stable Diffusion backbone via a lightweight cross-attention adapter. This ensures that the synthesized high-accuracy textures are based on absolutely accurate structural predictions. Evaluated on a global Sentinel-2 dataset, Sat-JEPA-Diff excels at resolving sharp boundaries. It achieves leading perceptual scores (GSSIM: 0.8984, FID: 0.1475) and significantly outperforms deterministic baselines, despite standard autoregressive stability limits. The code and dataset are publicly available on https://github.com/VU-AIML/SAT-JEPA-DIFF. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2603_13943 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | Sat-JEPA-Diff: Bridging Self-Supervised Learning and Generative Diffusion for Remote Sensing Komurcu, Kursat Petkevicius, Linas Computer Vision and Pattern Recognition Machine Learning 68T45, 68T07, 68U10 I.2.6; I.4.8; I.4.10; I.2.10 Predicting satellite imagery requires a balance between structural accuracy and textural detail. Standard deterministic methods like PredRNN or SimVP minimize pixel-based errors but suffer from the "regression to the mean" problem, producing blurry outputs that obscure subtle geographic-spatial features. Generative models provide realistic textures but often misleadingly reveal structural anomalies. To bridge this gap, we introduce Sat-JEPA-Diff, which combines Self-Supervised Learning (SSL) with Hidden Diffusion Models (LDM). An IJEPA module predicts stable semantic representations, which then route a frozen Stable Diffusion backbone via a lightweight cross-attention adapter. This ensures that the synthesized high-accuracy textures are based on absolutely accurate structural predictions. Evaluated on a global Sentinel-2 dataset, Sat-JEPA-Diff excels at resolving sharp boundaries. It achieves leading perceptual scores (GSSIM: 0.8984, FID: 0.1475) and significantly outperforms deterministic baselines, despite standard autoregressive stability limits. The code and dataset are publicly available on https://github.com/VU-AIML/SAT-JEPA-DIFF. |
| title | Sat-JEPA-Diff: Bridging Self-Supervised Learning and Generative Diffusion for Remote Sensing |
| topic | Computer Vision and Pattern Recognition Machine Learning 68T45, 68T07, 68U10 I.2.6; I.4.8; I.4.10; I.2.10 |
| url | https://arxiv.org/abs/2603.13943 |