Stable training of autoencoders for hyperspectral unmixing

TytułStable training of autoencoders for hyperspectral unmixing
Publication TypeJournal Article
Rok publikacjiSubmitted
AutorzyKsiążek K, Głomb P, Romaszewski M, Cholewa M, Grabowski B
JournalarXiv preprint 2109.13748
Date Published09/2021
Abstract

Neural networks, autoencoders in particular, are one of the most promising solutions for unmixing hyperspectral data, i.e. reconstructing the spectra of observed substances (endmembers) and their relative mixing fractions (abundances). Unmixing is needed for effective hyperspectral analysis and classification. However, as we show in this paper, the training of autoencoders for unmixing is highly dependent on weights initialisation. Some sets of weights lead to degenerate or low performance solutions, introducing negative bias in expected performance. In this work we present the results of experiments investigating autoencoders' stability, verifying the dependence of reconstruction error on initial weights and exploring conditions needed for successful optimisation of autoencoder parameters.

URLhttps://arxiv.org/abs/2109.13748