UnCRtainTS: Uncertainty Quantification for Cloud Removal in Optical Satellite Time Series

11 Apr 2023  ยท  Patrick Ebel, Vivien Sainte Fare Garnot, Michael Schmitt, Jan Dirk Wegner, Xiao Xiang Zhu ยท

Clouds and haze often occlude optical satellite images, hindering continuous, dense monitoring of the Earth's surface. Although modern deep learning methods can implicitly learn to ignore such occlusions, explicit cloud removal as pre-processing enables manual interpretation and allows training models when only few annotations are available. Cloud removal is challenging due to the wide range of occlusion scenarios -- from scenes partially visible through haze, to completely opaque cloud coverage. Furthermore, integrating reconstructed images in downstream applications would greatly benefit from trustworthy quality assessment. In this paper, we introduce UnCRtainTS, a method for multi-temporal cloud removal combining a novel attention-based architecture, and a formulation for multivariate uncertainty prediction. These two components combined set a new state-of-the-art performance in terms of image reconstruction on two public cloud removal datasets. Additionally, we show how the well-calibrated predicted uncertainties enable a precise control of the reconstruction quality.

PDF Abstract

Results from the Paper


Task Dataset Model Metric Name Metric Value Global Rank Result Benchmark
Cloud Removal SEN12MS-CR UnCRtainTS L2 MAE 0.027 # 1
PSNR 28.90 # 1
SAM 8.320 # 1
SSIM 0.880 # 2
Cloud Removal SEN12MS-CR-TS UnCRtainTS L2 RMSE 0.049 # 2
PSNR 27.23 # 3
SSIM 0.859 # 2
SAM 10.168 # 2
Cloud Removal SEN12MS-CR-TS UnCRtainTS ฯƒ RMSE 0.051 # 3
PSNR 27.84 # 2
SSIM 0.866 # 1
SAM 10.160 # 1

Methods


No methods listed for this paper. Add relevant methods here