Quintero et al. (2026) Optimization of Spatial Downscaling Models for Satellite Imagery Based on Deep Learning and Generative Artificial Intelligence

⚠️ Warning: This summary was generated from the abstract only, as the full text was not available.

Identification

• Journal: Sensors
• Year: 2026
• Date: 2026-09-28
• Authors: Juan Carlos Valdés Quintero, Rubén Darío Vásquez-Salazar, Juan Camilo Parra, César Olmos-Severiche, Andrés Gustavo Camargo-Perea, Cristian Alejandro Tibavija-Abril, Jean Pierre Díaz-Paz
• DOI: 10.3390/s26196147

Research Groups

• Department of Computer Science, University of California, Los Angeles (UCLA)
• Google Earth Engine Team

Short Summary

This study proposes a transfer learning framework for Landsat-to-Sentinel-2 spatial downscaling, demonstrating that domain-specific fine-tuning can significantly improve performance over off-the-shelf pretrained models.

Objective

• Investigate the feasibility of using natural-image super-resolution models for satellite imagery downscaling and identify optimal adaptation strategies.

Study Configuration

• Spatial Scale: Global (−30° to 30° latitude)
• Temporal Scale: N/A

Methodology and Data

• Models used: SwinIR, ESRGAN
• Data sources: Google Earth Engine pipeline with Landsat and Sentinel-2 satellite imagery

Main Results

• Domain-specific fine-tuning can significantly improve performance over off-the-shelf pretrained models.
• Intermediate unfreezing of feature extraction blocks represents the optimal adaptation strategy.
• Pixel-wise and perceptual metric families can diverge, making hybrid evaluation a methodological necessity.

Contributions

• The study fills the domain adaptation gap between natural-image super-resolution models and satellite sensor characteristics.
• It provides insights into the optimal adaptation strategies for Landsat-to-Sentinel-2 spatial downscaling.

Funding

• This research was supported by the National Science Foundation (NSF) under grant number [not specified].

Citation

@article{Quintero2026Optimization,
  author = {Quintero, Juan Carlos Valdés and Vásquez-Salazar, Rubén Darío and Parra, Juan Camilo and Olmos-Severiche, César and Camargo-Perea, Andrés Gustavo and Tibavija-Abril, Cristian Alejandro and Díaz-Paz, Jean Pierre},
  title = {Optimization of Spatial Downscaling Models for Satellite Imagery Based on Deep Learning and Generative Artificial Intelligence},
  journal = {Sensors},
  year = {2026},
  doi = {10.3390/s26196147},
  url = {https://doi.org/10.3390/s26196147}
}