SpectroscopyNet: Learning to pre-process Spectroscopy Signals without clean data

TL;DR

SpectroscopyNet introduces a deep learning method that effectively cleans spectroscopy signals without requiring clean training data, by disentangling signal and noise components using a siamese neural network.

Contribution

The paper presents a novel siamese neural network framework that learns to separate signal from noise in spectroscopy data without clean labels, improving cleaning performance over traditional methods.

Findings

Outperforms standard feature engineering approaches on LIBS data

Successfully disentangles signal and noise components

Enhances measurement quality for spectroscopy signals

Abstract

In this work we propose a deep learning approach to clean spectroscopy signals using only uncleaned data. Cleaning signals from spectroscopy instrument noise is challenging as noise exhibits an unknown, non-zero mean, multivariate distributions. Our framework is a siamese neural net that learns identifiable disentanglement of the signal and noise components under a stationarity assumption. The disentangled representations satisfy reconstruction fidelity, reduce consistencies with measurements of unrelated targets and imposes relaxed-orthogonality constraints between the signal and noise representations. Evaluations on a laser induced breakdown spectroscopy (LIBS) dataset from the ChemCam instrument onboard the Martian Curiosity rover show a superior performance in cleaning LIBS measurements compared to the standard feature engineered approaches being used by the ChemCam team.