Fig. 4: Performance comparison using various metrics.
From: A machine learning model for textured X-ray scattering and diffraction image denoising
Performance of networks under optimal hyperparameter configurations using various metrics, including a RMSE, b Betas, c PSNR, d Delta Direct Angle, e SSIM, f Delta Mean, g Delta Amplitude, h Delta Standard Deviation. Two variants of SEDCNN, PMRID, and REDCNN were compared. Units: RMSE, PSNR, SSIM, Betas are in arbitrary units; Delta Amplitude is in image intensity values; Delta Direct Angle, Delta Mean, Delta Standard Deviation are in degrees. The number of epochs was set to 500.
