IEEE Transactions on Geoscience and Remote Sensing (TGRS)

Kernel-Adaptive
Optimization

A diffusion framework for high-resolution satellite image inpainting across urban, forest, and agricultural landscapes. KAO introduces adaptive kernel-aware optimization for structure-preserving geospatial reconstruction and achieves superior fidelity over existing diffusion baselines.

Teerapong Panboonyuen Β· Chulalongkorn University
KAO Teaser
Journal
IEEE TGRS
Impact Factor
8.6
Task
Satellite AI
Presented At
AOGS 2026

Overview

KAO combines diffusion-based generative modeling with Kernel-Adaptive Optimization and Token Pyramid Transformers to reconstruct missing satellite regions with high spatial fidelity. The framework is specifically designed for Earth observation imagery where structural continuity and texture realism are critical.

Published in IEEE TGRS

IEEE Transactions on Geoscience and Remote Sensing β€” one of the world’s leading journals in remote sensing and geospatial AI.

IF 8.6

🌏 Invited Presentation at AOGS 2026

This work has been invited for presentation at the 23rd Annual Meeting of the Asia Oceania Geosciences Society (AOGS 2026), showcasing generative AI for high-resolution satellite image inpainting across diverse geospatial environments.

View Official AOGS Program

Urban Reconstruction

Preserves roads, buildings, and fine-grained urban topology under large occlusions and missing regions.

Forest & Agriculture

Maintains vegetation consistency, spatial continuity, and landscape textures across rural environments.

Adaptive Diffusion

Dynamically modulates optimization focus using kernel-aware weighting strategies inside latent diffusion space.

Qualitative Results

KAO consistently generates sharper, more realistic, and semantically coherent reconstructions than Stable Diffusion, RePaint, SatDiff, DPS, and PSLD across challenging remote sensing scenarios.

Main Results
Comparison across multiple diffusion baselines. KAO restores structural layouts and textures with superior fidelity.
Scene 1
Scene 2
Scene 3
Scene 4
Scene 5
Scene 6

Technical Highlights

Kernel-Adaptive Optimization

Spatially adaptive weighting allows the diffusion process to focus on difficult reconstruction regions and semantic inconsistencies.

Diffusion Prior Modeling

Robust latent diffusion sampling improves realism and preserves geospatial continuity under severe missing data conditions.

Remote Sensing AI

Designed for modern geospatial foundation models, Earth observation systems, and large-scale satellite intelligence.

BibTeX

@article{panboonyuen2025kao,
  author    = {Teerapong Panboonyuen},
  title     = {KAO: Kernel-Adaptive Optimization in Diffusion for Satellite Image Inpainting},
  journal   = {IEEE Transactions on Geoscience and Remote Sensing},
  year      = {2025},
  doi       = {10.1109/TGRS.2025.3621738},
  publisher = {IEEE}
}