Yuan et al. (2026) CosmDiff: Integrating Multitemporal Optical-SAR Data With Conditional Diffusion Models for Optical Satellite Time Series Reconstruction

Identification

• Journal: IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing
• Year: 2026
• Date: 2026-01-01
• Authors: Yuan Yuan, Junhan Zhou, Lei Lin, Ying Yu, Qingshan Liu
• DOI: 10.1109/jstars.2026.3655691

Research Groups

[]

Short Summary

This paper introduces CosmDiff, a method for reconstructing optical satellite time series by integrating multitemporal optical and Synthetic Aperture Radar (SAR) data using conditional diffusion models.

Objective

• To reconstruct optical satellite time series by integrating multitemporal optical-SAR data with conditional diffusion models.

Study Configuration

• Spatial Scale: []
• Temporal Scale: Multitemporal

Methodology and Data

• Models used: Conditional Diffusion Models (specifically "CosmDiff")
• Data sources: Multitemporal Optical-SAR data (implied satellite observations)

Main Results

[]

Contributions

[]

Funding

[]

Citation

@article{Yuan2026CosmDiff,
  author = {Yuan, Yuan and Zhou, Junhan and Lin, Lei and Yu, Ying and Liu, Qingshan},
  title = {CosmDiff: Integrating Multitemporal Optical-SAR Data With Conditional Diffusion Models for Optical Satellite Time Series Reconstruction},
  journal = {IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing},
  year = {2026},
  doi = {10.1109/jstars.2026.3655691},
  url = {https://doi.org/10.1109/jstars.2026.3655691}
}