Mobility & logisticsResearch findings
HiPredict: Machine Learning to Reconstruct Missing HiRISE RED4 Channel
Original published 2026.10.16 20:40 KSTAdded to Blackdeck 2026.10.19 11:32 KST
What has changed?
Uses non-linear spatiotemporal-spectral relationships to reconstruct missing RED4 channel more accurately than linear methods
Key points
- Proposes machine learning solution for missing RED4 channel due to detector failure in HiRISE
- Leverages U-Net architecture to learn non-linear relationships from IR10, BG12, and metadata
- Achieves 1.1% normalized mean absolute error, significantly outperforming linear synthetic method (3.1%)
Connect it to your work
Introduces a machine learning approach to recover data loss caused by detector failure in Mars orbiting mission
What remains unproven
This research note is based on an abstract; check the original for test conditions and application scope.
Originals used for this note
Links open the original publisher in a new tab.