Fig. 4: RIXS spectra recorded on SrTiO3.
From: Weak signal extraction enabled by deep neural network denoising of diffraction data
a–c, Top row: RIXS spectra with counting statistics of 1 (a), 4 (b) and 40 (c) times 3 min (t). Left: counting intensities as detector channel versus energy loss. Right: horizontally projected RIXS spectra. Bottom row: corresponding denoised neural network outputs. The arrows in c highlight three inelastic peaks.
