Spectroscopic data de-noising via training-set-free deep learning method

De-noising plays a crucial role in the post-processing of spectra. Machine learning-based methods show good performance extracting intrinsic information from noisy data, but often require high-quality training set that is typically inaccessible real experimental measurements. Here, using spectra angle-resolved photoemission spectroscopy (ARPES) as an example, we develop de-noising method for spectral without need set. This possible our leverages self-correlation themselves. It preserves energy band features and thus facilitates further analysis processing. Moreover, since not limited by specific properties compared to previous ones, it may well be extended other fields application scenarios where obtaining multidimensional data challenging.