| Title |
Sat-JEPA-Diff: bridging self-supervised learning and generative diffusion for remote sensing |
| Authors |
Komurcu, Kursat ; Petkevičius, Linas |
| Full Text |
|
| Is Part of |
4th ICLR workshop on machine learning for remote sensing, Rio de Janeiro, Brazil.. Rio de Janeiro. 2026, p. [1-10] |
| Keywords [eng] |
Remote sensing ; I-JEPA ; generative AI |
| Abstract [eng] |
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 geographicspatial 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 Latent 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 GitHub. |
| Published |
Rio de Janeiro |
| Type |
Conference paper |
| Language |
English |
| Publication date |
2026 |
| CC license |
|
