Uncertainty Quantification with Deep Ensemble methods for Super-Resolution of Sentinel 2 Satellite Images

Abstract

The recently deployed Sentinel-2 satellite constellation produces images in 13 wavelength bands with a Ground Sampling Distance (GSD) of 10m, 20m and 60m. Super-resolution aims to generate all 13 bands with a spatial resolution of 10m. This paper investigates the performance of DSen2, a proposed convolutional neural network (CNN) based method, to tackle super-resolution, in terms of accuracy and uncertainty. As the optimization problem for obtaining the weights of a CNN is highly non-convex, there are multiple different local minima for the loss function. This results in several possible CNN models with different weights and, thus, implies epistemic ncertainty. In this work methods to quantify epistemic uncertainty, coined weighted deep ensembles (WDESen2) and its variants, are proposed. It allows to quantify predictive uncertainty estimates and, moreover, to improve the accuracy of the prediction by selective prediction. It consists of considering deep ensembles, where each model’s importance can be weighted depending on the model’s validation loss. We show that weighted deep ensembles improve the accuracy of the prediction, compared to state of the art methods and deep ensembles. Moreover, the uncertainties can be linked to the underlying inverse problem and physical patterns on the ground. This allows to improve the trustworthiness of CNN predictions and the predictive accuracy with selective prediction