Raw data underlying the publication: A Lightweight Deep Learning Framework for Fast, Real-Time Super-Resolution Fluctuation Imaging
What's in it
RESURF DATA REPOSITORY: This repository provides data which is used for our paper: A Lightweight Deep Learning Framework for Fast, Real-Time Super-Resolution Fluctuation Imaging (RESURF). RESURF is a deep-learning based real-time SR imaging method. We used a recurrent neural network based model ( MISRGRU) and trained it using realistic fluorescence microscopy simulations. We simulated a large dataset containing imaging conditions of different SNRs and blinking conditions and used it to create multiple foundation models. We fine-tuned foundation models with smaller datasets ( either experimental or simulated). In this repository, we provide the complete dataset we used for training and testing as well as foundation models and fine-tuned models. 1- Foundation Models, created by using large synthetic dataset (FWHM:2.2 pixels)2- Live-cell Models, used in Figure 3-4-5-6 (FWHM:2.7 pixels)3- Simulations Foundation_models_testdata: Synthetic test set used for testing the foundation models (Figure 1-2) Live_cell_simulations: Simulations with high-background imitating fluorescence protein conditions which is used for training live-cell models (transfer learning from foundation models)4- …
Files and services
- DataCite REST API (JSON)api.datacite.orgOpen