Automatic Baseline Correction of 1D Signals Using a Parameter-Free Deep Convolutional Autoencoder Algorithm

Le résumé fourni par la source

Baseline correction techniques are highly applicable in analytical chemistry. Consequently, there is a constant demand for universal and automated baseline correction methods. Our new procedure based on the Convolutional Autoencoder model (ConvAuto model) combined with an automated algorithm to Apply the Model (ApplyModel procedure) meets those expectations. The key advantage of this idea is its ability to handle 1D signals of various lengths and resolutions, which is a common limitation encountered for deep neural network models. The proposed procedure is fully automatic and does not require any parameter optimization. As our experiments have shown, the ApplyModel procedure can also be easily combined with other baseline correction methods that utilize deep neural networks, such as the ResUNet model, which extended it practical application too. The usability of our new approach was tested by its implementation for both simulated and experimental signals, ranging from 200 to 4000 point length. The results were compared with those provided by the ResUNet model.

Ce résumé expose les affirmations des auteurs. BNTIC ne l’interprète pas comme une validation indépendante des résultats.