Image restoration for optical synthetic aperture system via variational physics-informed network

A three-stage deep learning-based training frame for spectra baseline correction

“…Recently, we have witnessed the rapid development of deep learning in image processing. [22][23][24] However, the progress of deep learning-based baseline correction methods is relatively slower. Therefore, what makes deep learning in spectra baseline correction lag behind image processing?…”