A CycleGAN that converts EMIT satellite methane retrievals into airborne-style images, followed by an airborne-trained classifier, achieves F1 0.88 and outperforms zero-shot, satellite-only, and fine-tuned models.
Toward global mapping of methane with TROPOMI: First results and intersatellite comparison to GOSAT,
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
citation-role summary
background 1
citation-polarity summary
fields
eess.SP 1years
2025 1verdicts
CONDITIONAL 1roles
background 1polarities
background 1representative citing papers
citing papers explorer
-
Multi-Platform Methane Plume Detection via Model and Domain Adaptation
A CycleGAN that converts EMIT satellite methane retrievals into airborne-style images, followed by an airborne-trained classifier, achieves F1 0.88 and outperforms zero-shot, satellite-only, and fine-tuned models.