Baseline Correction and Denoising of Raman Spectra by Deep Residual CNN

A practical deep residual CNN model for the improvement of Raman spectral analysis is proposed in this work. Through the feature transformation in the deep residual blocks, it successfully denoises and precisely corrects the baseline.