Subject: Data for 3D RCAN Paper Release date: xx-xx-2021 (MM-DD-YYYY) Contacts: Jiji Chen (jiji.chen@nih.gov), Hari Shroff (hari.shroff@nih.gov) Organization: Advanced Imaging and Microscopy (AIM) Resource, National Institutes of Health This data release contains all training and testing data for the 3D RCAN paper published in Nature Methods: Three-dimensional residual channel attention networks denoise and sharpen fluorescence microscopy image volumes Data may be found at: Data size: After upzip, the total size is 521 GB. Please leave enough space in the disk before unzip. Data are organized in four categories: - Denoising: It contains all training and testing data for Actin, ER, Golgi, lysosome, Microtubule, Tomm20 Mitochondria and Matrix protein Mitochondria. It also contains one live cell testing data (U2OS cells transfected with maxtrix protein Mitochondria.) - Phantom Spheres: It contains training and testing data for synthetic blurred phantom sphere(2, 3 and 4 times). The ground truth is the synthetic phantom sphere without blurring. - Confocal to STED: It contains training and testing data for three different structure shown in the paper (Microtubule, nuclear pore complex and DNA). Raw and GT data in the training and testing folder refer to confocal and STED images. - Expansion Microscopy: It contains training and testing data for two different structure shown in the paper. (Microtubule and Tomm20 stained Mitochondria). Raw and GT data in the training and testing folder refer to Synthetic raw and deconvolved expansion microscopy iSIM images. 3D RCAN code is available on our Github repo: https://github.com/AiviaCommunity/3D-RCAN Notes: 1. Gt refers to Ground truth. Raw refers to the raw data. 2. STED Stimulated emission depletion microscopy 3. iSIM: instant structured illumination microscopy License: Licensed under BSD 3-Clause License.