Proposes ISensD and ESensI methods to increase robustness of multi-sensor EO models to missing sensors, with experiments on three temporal datasets showing ensemble models are most robust.
arXiv preprint arXiv:2401.07782 (2024) 12 F
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.LG 1years
2024 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
Increasing the Robustness of Model Predictions to Missing Sensors in Earth Observation
Proposes ISensD and ESensI methods to increase robustness of multi-sensor EO models to missing sensors, with experiments on three temporal datasets showing ensemble models are most robust.