Pseudo-labelled uncertainty quantification is used a lot in imagery, but I want to use my time series data to generate reliable and unreliable data from pseudo-labelled methods, such as "UnCRtainTS: Uncertainty Quantification for Cloud Removal in Optical Satellite Time Series" this paper, how can I do that please?
Trying to use this method, UnCRtainTS, to replace the image with my time series data on the input side to come up with reliable and unreliable data! The language is python.