Find research datasets worth reusing

Search datasets from major research repositories and use ShareScore to quickly assess how well each record supports discovery, access, and reuse.

datasets available to search

ShareScore release 0.9.0

Dataset results

23 results for “Generative Adversarial Network”

Learn how ShareScore rates datasets ↗

zenodo52/100

Data archive for "Stochastic Super-Resolution for Downscaling Time-Evolving Atmospheric Fields with a Generative Adversarial Network"

This datasets supports the paper "Stochastic Super-Resolution for Downscaling Time-Evolving Atmospheric Fields with a Generative Adversarial Network" submitted to IEEE Transactions in Geoscience and Remote Sensing. A preprint of the paper can be found here: <a href="https://arxiv.org/abs/2005.10374">https://arxiv.org/abs/2005.10374</a>. The code that uses these data is available at <a href="https://github.com/jleinonen/downscaling-rnn-gan">https://github.com/jleinonen/downscaling-rnn-gan</a>. The file "goes-samples-2019-128x128.nc" contains the training dataset called "GOES-COT" in the paper, consisting of cloud optical depth measurements from the GOES-16 satellite. The files "gen_weights*.nc" contain the generator weights saved at different time steps during training for the two different datasets described in the paper.

opencc-by-4.0May 2020View details →

zenodo48/100

Pythia Generated Jet Images with Alternative Rotation Scheme for Location Aware Generative Adversarial Network Training

Dataset containing 300k jet images that can be used to train Location Aware Generative Adversarial Networks (LAGAN) for High Energy Physics, such as the one in [arXiv:1701.05927]. Format: HDF5 file with the following fields: <ul> <li>'image' : array of dim (300000, 25, 25), contains the pixel intensities of each 25x25 image</li> <li>'signal' : binary array to identify signal (1, i.e. W boson) vs background (0, i.e. QCD)</li> <li>'jet_eta': eta coordinate per jet</li> <li>'jet_phi': phi coordinate per jet</li> <li>'jet_mass': mass per jet</li> <li>'jet_pt': transverse momentum per jet</li> <li>'jet_delta_R': distance between leading and subleading subjets if 2 subjets present, else 0</li> <li>'tau_1', 'tau_2', 'tau_3': substructure variables per jet (a.k.a. n-subjettiness, where n=1, 2, 3)</li> <li>'tau_21': tau2/tau1 per jet</li> <li>'tau_32': tau3/tau2 per jet</li> </ul> Details: <ul> <li>Simulated using Pythia 8.219 at √ s = 14 TeV</li> <li>Image pre-processing using method from in L. de Oliveira et al., Jet-Images -- Deep Learning Edition [arXiv:1511.05190]</li> <li>scikit-image==0.10.0 implementation of cubic spline rotation with fewer low energy artifacts than scikit-image>=0.12.0</li> <li>Finite calorimeter granularity simulated with 0.1×0.1 grid in η and φ, with η × φ ∈ [−1.25, 1.25] × [−1.25, 1.25]</li> <li>Jet clustering with anti-kt algorithm with a radius R = 1.0 using FastJet 3.2.1; constituent re-clustering into R = 0.3 kt subjets</li> <li>Intensity of pixel = pT of cell</li> <li>60 GeV < mjet < 100 GeV</li> <li>250 GeV < pTjet < 300 GeV</li> <li>Sparse images (~10% NNZ)</li> </ul> Full dataset description in [arXiv:1701.05927].

opencc-by-4.0Feb 2017View details →

zenodo40/100

A Robust Generative Adversarial Network Approach for Climate Downscaling and Weather Generation

<h1>Dataset Description for "A Robust Generative Adversarial Network Approach for Climate Downscaling and Weather Generation"</h1> This dataset accompanies the research paper titled "A Robust Generative Adversarial Network Approach for Climate Downscaling and Weather Generation", currently under review for the AGU Journal JAMES. The study introduces a novel Regional Climate Model (RCM) emulator focusing on high-resolution climate downscaling for the New Zealand region. For additional insights and access to the codebase utilized in this research, please refer to our <a href="https://github.com/nram812/A-Robust-Generative-Adversarial-Network-Approach-for-Climate-Downscaling" target="_new">GitHub repository</a>. <h2>Aims</h2> Our study's overarching goal was to assess the effectiveness of Generative Adversarial Networks (GANs) in a climate downscaling context and is structured around two aims. The first aim of our study is to examine whether GANs can overcome several important limitations of regression-based climate downscaling algorithms (i.e. underestimating the magnitude of extreme events). The second and most important aim of our study is to assess the robustness GAN performance to different training hyperparameters. Our robustness assessment thoroughly scrutinizes GANs for their application in climate downscaling contexts, ensuring that they can learn and capture regional climate processes <h2>Geographic Focus</h2> Our research focuses only on the New Zealand Region (165°E-184°W, 33°S-51°S).   <h2>Data Overview</h2> <h3>Training and Evaluation Data</h3> The training data used in this study (for our RCM emulator) only spans the historical period of simulation. It comprises daily accumulated precipitation as the primary target variable, alongside large-scale predictor variables.  <ul> <li> Resolution: The target variable is presented at a 12km resolution, reflecting the highest resolution face of RCM for the New Zealand region. Predictor variables are coarsened to a 1.5-degree resolution from original CCAM outputs using conservative interpolation.  </li> <li> Period Coverage: <ul> <li>Training Data: 1960-2014</li> <li>Validation Data: 1986-2005</li> </ul> </li> <li> Models: <ul> <li>Training on: ACCESS-CM2</li> <li>Validated on: EC-Earth3, NorESM2-MM</li> </ul> </li> </ul> <h3>File Structure</h3> <ul> <li> Training Data: <ul> <li>Target/Ground Truth (Y): <code>predictor_ACCESS-CM2_hist.nc</code></li> <li>Predictor (X): <code>pr_ACCESS-CM2_hist.nc</code></li> </ul> </li> <li> Evaluation Data: <ul> <li>NorESM2-MM: <ul> <li>Target (Y): <code>NorESM2-MM_historical_precip_compressed.nc</code></li> <li>Predictor (X): <code>NorESM2-MM_histupdated_compressed.nc</code></li> </ul> </li> <li>EC-Earth3: <ul> <li>Target: <code>EC-Earth3_historical_precip_compressed.nc</code></li> <li>Predictor: <code>EC-Earth3_histupdated_compressed.nc</code></li> </ul> </li> </ul> </li> </ul> <h2>Methodological Insights</h2> <ul> <li> Regional Climate Model, Our Regional Climate Model training data is from the Conformal Cubic Atmospheric Model (CCAM) which is a global non-hydrostatic atmospheric model renowned for its variable-resolution cubic grid. . For more information about CCAM, please see the following <a href="https://agupubs.onlinelibrary.wiley.com/doi/abs/10.1029/2023JD038530">paper</a>. </li> <li> Predictor and Target Variables: Daily-averaged large-scale prognostic variables, including zonal wind, meridional wind, temperature, and specific humidity, are employed as predictors at the 500mb and 850mb pressure levels. These are normalized (see the GitHub repository for the mean and standard deviation fields). Precipitation is taken as is from CCAM and accumulated for each given day. Static predictors are also used in our model, which is stored in a GitHub repository. </li> <li> Training Framework: Our dataset benefits from the "perfect framework" training strategy, which uses CCAM-coarsened predictor variables. For more information about the perfect and imperfect training frameworks, see the following <a title="review" href="https://journals.ametsoc.org/view/journals/aies/3/2/AIES-D-23-0066.1.xml">review</a> </li> </ul>

opencc-by-4.0Apr 2024View details →

zenodo40/100

Supplementary files for Reconstructing Kinetic Models for Dynamical Studies of Metabolism using Generative Adversarial Networks, main part

Supplementary files containing datasets needed to reproduce the results of the manuscript "Reconstructing Kinetic Models for Dynamical Studies of Metabolism using Generative Adversarial Networks" by S. Choudhury et al. The code to use with these data and reproduce the manuscript results is available at  <a href="https://github.com/EPFL-LCSB/rekindle">https://github.com/EPFL-LCSB/rekindle</a> and <a href="https://gitlab.com/EPFL-LCSB/rekindle">https://gitlab.com/EPFL-LCSB/rekindle</a>. The execution of parts of this code is dependent on the SkimPy toolbox (<a href="https://github.com/EPFL-LCSB/skimpy">https://github.com/EPFL-LCSB/skimpy</a>). Refer to the readme files on the REKINDLE code repositories for more details. Datasets: <ul> <li> models.zip - Datasets parameterizing kinetic nonlinear models of a wild-type E. coli strain used for training generative adversarial networks <ul> <li>subfolder 1: kinetic - contains the kinetic model (kin_varma_curated.yml)</li> <li>subfolder 2:  thermo - contains the thermodynamic model for all the four physiologies (varma_fdp1, varma_fdp2, varma_fdp3, varma_fdp4)</li> <li>subfolder 3:  steady_state_samples: contains the TFA steady state profiles for all four physiologies (samples_fdp1, sample_fdp2, samples_fdp3, samples_fdp4)</li> <li>subfolder 4: parameters - contains the kinetic parameter training dataset for each physiology (.hdf5 files), maximal eigenvalues (training labels)  (maximal_eigenvalues.csv) and the minimum eigenvalues (minimal_eigenvalues.csv)</li> </ul> </li> <li>vanilla_learning_training.zip: contains 4 folders for each of the 4 physiologies. <ul> <li>each of these folders contains 6 subsubfolders in the format N-{n} ( N-10, N-50, N-100, N-500, N-1000, N-72000), where {n} represents the number of used training data samples.</li> <li>every subsubfolder N-{n} contains 5 repeats folders. Each repeat folder contains, <ul> <li>E_-1.npy - GAN generated kinetic parameters at E-th epoch/</li> <li>E_-1_max_eig.csv - the maximal eigenvalues of Jacobian for E_-1.npy (Note: eigenvalues were not calculated for N=10, 50, 100 as traning failed)/</li> </ul> </li> </ul> </li> <li>transfer_learning_training.zip - contains 12 subfolders "tl_fdpi_fdpj" where i,j ={1,2,3,4} for each of the 12 transfer learning case <ul> <li>each of these folders contains 5 subsubfolders N-10, N-50, N-100, N-500, N-1000</li> <li>every subsubfolder N-{n} contains 5 repeats folders. Each repeat folder contains, <ul> <li>E_-1.npy - GAN generated kinetic parameters at E-th epoch/</li> <li>E_-1_max_eig.csv - the maximal eigenvalues of Jacobian for E_-1.npy </li> </ul> </li> </ul> </li> </ul> <ul> <li>best_generators.zip <ul> <li>The best generators (with the highest incidence of relevant models) for each physiology (generator1- 4.h5)</li> <li>The normalizing scaling parameters for each generator (d_scaling.pkl).</li> <li> </li> </ul> </li> <li>Temporal evolution of perturbations in non-linear ordinary differential equations <ul> <li>vanilla_ODE_sample_parameters.zip - contains (i) 1000 REKINDLE generated kinetic parameter sets for each of the 4 physiologies and their corresponding eigenvalues (4 in total) (ii) 1000 ORACLE generated kinetic parameter sets for each of the 4 physiologies and their corresponding eigenvalues (4 in total). These parameter sets parameterize the ODEs which are integrated.</li> <li>ode_solutions_physiology1.zip (available at <a href="https://zenodo.org/record/5818192">https://zenodo.org/record/5818192</a>) -  contains 100 subfolders, each subfolder containing the time-series evolution data of 1000 kinetic models parameterized by REKINDLE generated parameter sets for physiology 1, each of the 1000 models having a random perturbation.</li> <li>ode_solutions_physiology1_ORACLE.zip (available at <a href="https://zenodo.org/record/5819669">https://zenodo.org/record/5819669</a>) -  contains 100 subfolders, each subfolder containing the time-series evolution data of 1000 kinetic models parameterized by ORACLE generated parameter sets for physiology 1, each of the 1000 models having a random perturbation.</li> <li>ode_solutions_physiologies2-4.zip - contains 6 subfolders (physiology_2-4, physiology_2-4_ORACLE), with each subfolder containing 10 sub subfolders. Each sub subfolder containing the time-series evolution data of 1000 kinetic models parameterized by REKINDLE / ORACLE generated parameter sets for physiology 2-4, each of the 1000 models having a random perturbation.</li> <li>transfer_learning_ODE_solutions.zip - contains two subfolders N_10, N_50, each subfolder contains 12 subsubfolders titled i_j (where i = {1,2,3,4} and j = {1,2,3,4} where 1_2 represent the transfer learning case from physiology 2 to physiology 1 and when using {n} samples from physiology 2 and so on (where {n}=10 and 50 respectively).  Each subsubfolders contain <ul> <li>i_j.hdf5: contains 300 kinetic parameter sets generated using (i) REKINDLE for this transfer learning case</li> <li>i_j.csv: the maximal eigenvalues of the parameter sets</li> <li>solutions.csv: ODE integrated time series data for the relevant kinetic parameters out of the 300 generated.</li> </ul> </li> </ul> </li> </ul>

opencc-by-4.0Jan 2022View details →

zenodo40/100

Supplementary files for Reconstructing Kinetic Models for Dynamical Studies of Metabolism using Generative Adversarial Networks, additional part 2

opencc-by-4.0Jan 2022View details →

zenodo40/100

Supplementary files for Reconstructing Kinetic Models for Dynamical Studies of Metabolism using Generative Adversarial Networks, additional part 1

opencc-by-4.0Jan 2022View details →

zenodo40/100

Image Synthesis with a Convolutional Capsule Generative Adversarial Network- Dataset

Dataset of 152 two-photon images (512x512) of axons with segmentation labels.  We combined data from two published sources (Bass et al., 2017; Canty et al., 2018) to get 152 (512×512) 2D images (produced from 3D image stacks), and manually produced the corresponding labels. These images were collected using in-vivo two-photon microscopy from the mouse somatosensory cortex.  To generate the 2D images, we used a max projection over the 3D stack. The labels are binary segmentation maps of the axons. This dataset is split into a train (132 images) and test (20 images) sets. The raw 2D images of axons are in /original folder, and the segmentation labels are in /mask folder. Please cite the following paper when using this dataset: Bass, C., Dai, T., Billot, B.,  Arulkumaran, K., Creswell, A., Clopath, C., De Paola, V., and Bharath, A. A., 2019. “Image synthesis with a convolutional capsule generative adversarial network,” Medial Imaging with Deep Learning. This dataset was complied from the following publications: Bass, C., Helkkula, P., De Paola, V., Clopath, C. and Bharath, A.A., 2017. Detection of axonal synapses in 3D two-photon images. PloS one, 12(9), p.e0183309. Canty, A.J., Jackson, J.S., Huang, L., Trabalza, A., Bass, C., Little, G. and De Paola, V., 2018. Single-axon-resolution intravital imaging reveals a rapid onset form of Wallerian degeneration in the adult neocortex. bioRxiv, p.391425.

opencc-by-4.0Feb 2019View details →

zenodo40/100

Dataset for IJCAI 2019 paper, Scribble-to-Painting Transformation with Multi-Task Generative Adversarial Networks

The dataset and pre-trained model for IJCAI 2019 paper "Scribble-to-Painting Transformation with Multi-Task Generative Adversarial Networks" <ul> <li>Pre-trained model(Pytorch): DSPNet_G_200_epochs.pth</li> <li>Dataset (generated from coco dataset): starry_night_coco.zip</li> <li>Github: https://github.com/jinningli/DSP-Net</li> </ul> Please cite our paper if you are using this dataset: Jinning Li, and Yexiang Xue. Scribble-to-Painting Transformation with Multi-Task Generative Adversarial Networks. In International Joint Conference on Artificial Intelligence (IJCAI) 2019 Abstract: We propose the Dual Scribble-to-Painting Network (DSP-Net), which is able to produce artistic paintings based on user-generated scribbles. In scribble-to-painting transformation, a neural net has to infer additional details of the image, given relatively sparse information contained in the outlines of the scribble. Therefore, it is more challenging than classical image style transfer, in which the information content is reduced from photos to paintings. Inspired by the human cognitive process, we propose a multi-task generative adversarial network, which consists of two jointly trained neural nets -- one for generating artistic images and the other one for semantic segmentation. We demonstrate that joint training on these two tasks brings in additional benefit. Experimental result shows that DSP-Net outperforms state-of-the-art models both visually and quantitatively. In addition, we publish a large dataset for scribble-to-painting transformation.

opencc-by-4.0Sep 2019View details →

zenodo40/100

Dataset for publication: Statistically Equivalent Virtual Microstructures for Modeling of Complex Polycrystalline Alloys Using a Generative Adversarial Network (GAN)-Enabled Computational Platform

This dataset provides the necessary data to get the images and results shown in the paper "Statistically Equivalent Virtual Microstructures for Modeling of Complex Polycrystalline Alloys Using a Generative Adversarial Network (GAN)-Enabled Computational Platform".  Source Data Raw.zip has the entire data set used to generate the images. Source Data.zip contains the processed data  from "Source Data Raw.zip".    Files with extension .dream3d are accompained by a file with extension .xdmf. This files can be opened with Paraview. And their data can be accesible using python or matlab. For more information contact Proffesor Somnath Ghosh at Johns Hopkins University, Civil and Systems Engineering Department.

opencc-by-4.0Aug 2024View details →

zenodo40/100

Using generative adversarial networks to match experimental and simulated inelastic neutron scattering data

Files uploaded here are related to the paper titled "Using generative adversarial networks to match experimental and simulated inelastic neutron scattering data". Here we investigate how generative adversarial networks can be used to match simulated- and experimental INS data.

opencc-by-4.0Nov 2022View details →

zenodo36/100

Pythia Generated Jet Images for Location Aware Generative Adversarial Network Training

Dataset containing 872666 jet images to train Location Aware Generative Adversarial Networks (LAGAN) for High Energy Physics. Results are published in [arXiv:1701.05927]. Format: HDF5 file with the following fields: <ul> <li>'image' : array of dim (872666, 25, 25), contains the pixel intensities of each 25x25 image</li> <li>'signal' : binary array to identify signal (1, i.e. W boson) vs background (0, i.e. QCD)</li> <li>'jet_eta': eta coordinate per jet</li> <li>'jet_phi': phi coordinate per jet</li> <li>'jet_mass': mass per jet</li> <li>'jet_pt': transverse momentum per jet</li> <li>'jet_delta_R': distance between leading and subleading subjets if 2 subjets present, else 0</li> <li>'tau_1', 'tau_2', 'tau_3': substructure variables per jet (a.k.a. n-subjettiness, where n=1, 2, 3)</li> <li>'tau_21': tau2/tau1 per jet</li> <li>'tau_32': tau3/tau2 per jet</li> </ul> Details: <ul> <li>Simulated using Pythia 8.219 at √ s = 14 TeV</li> <li>Image pre-processing using method from in L. de Oliveira et al., Jet-Images -- Deep Learning Edition [arXiv:1511.05190]</li> <li>scikit-image==0.12.0 implementation of cubic spline rotation</li> <li>Finite calorimeter granularity simulated with 0.1×0.1 grid in η and φ, with η × φ ∈ [−1.25, 1.25] × [−1.25, 1.25]</li> <li>Jet clustering with anti-kt algorithm with a radius R = 1.0 using FastJet 3.2.1; constituent re-clustering into R = 0.3 kt subjets</li> <li>Intensity of pixel = pT of cell</li> <li>60 GeV < mjet < 100 GeV</li> <li>250 GeV < pTjet < 300 GeV</li> <li>Sparse images (~10% NNZ)</li> </ul> Full dataset description in [arXiv:1701.05927].

opencc-by-4.0Dec 2016View details →

zenodo36/100

Dataset of "3D generative adversarial networks for turbulent flow estimation from wall measurements"

Dataset of the article 'Three-dimensional generative adversarial networks for turbulent flow estimation from wall measurements' (https://doi.org/10.1017/jfm.2024.432). The codes processing data here are on https://github.com/erc-nextflow/3D-GAN. This project has received funding from the European Research Council (ERC) under the European Union’s Horizon 2020 research and innovation programme (grant agreement no. 949085, NEXTFLOW). Views and opinions expressed are, however, those of the authors only, and do not necessarily reflect those of the European Union or the ERC. Neither the European Union nor the granting authority can be held responsible for them. A.C.M. acknowledges financial support from the Spanish Ministry of Universities under the Formación de Profesorado Universitario (FPU) programme 2020. R.V. acknowledges financial support from ERC (grant agreement no. 2021-CoG-101043998, DEEPCONTROL).

opencc-by-4.0Apr 2024View details →

zenodo36/100

Dataset for "On the Extrapolation of Generative Adversarial Networks for downscaling precipitation extremes in warmer climates"

<h1>Code and Dataset for "On the Extrapolation of Generative Adversarial Networks for downscaling precipitation extremes in warmer climates"</h1> This dataset accompanies the research paper titled "On the Extrapolation of Generative Adversarial Networks for downscaling precipitation extremes in warmer climates", currently under review for the AGU Journal GRL. The study introduces a novel Regional Climate Model (RCM) emulator focusing on high-resolution climate downscaling for the New Zealand region. For additional insights and access to the codebase utilized in this research, please refer to our <a href="https://github.com/nram812/On-the-Extrapolation-of-Generative-Adversarial-Networks-for-downscaling-precipitation-extremes">Github Repository</a>. The code can also be found as a ".zip" file: *On-the-Extrapolation-of-Generative-Adversarial-Networks-for-downscaling-precipitation-extremes-main.  <h2>Aims</h2> Our study focuses on two important gaps in the literature regarding the extrapolation of empirical downscaling algorithms. First, we examine how well relationships learned from a historical period extrapolate to future unobserved climates. We compare two widely used algorithms, a GAN and a deterministic CNN baseline, that use a similar architecture (i.e. convolutional layers) trained in a model-as-truth framework to downscale daily precipitation over New Zealand. We evaluate their accuracy in capturing climate change signals in mean and extreme precipitation. Second, we explore whether training on future vs. only historical periods combined with different-sized training datasets can improve extrapolation skill.  <h2>Geographic Focus</h2> Our research focuses only on the New Zealand Region (165°E-184°W, 33°S-51°S).   <h2>Data Overview</h2> <h3>Training and Evaluation Data</h3> The training data used in this study (for our RCM emulator) spans the historical period and future period (SSP370) of simulation. It comprises daily accumulated precipitation as the primary target variable, alongside large-scale predictor variables.  <ul> <li> Resolution: The target variable is presented at a 12km resolution, reflecting the highest resolution face of RCM for the New Zealand region. Predictor variables are coarsened to a 1.5-degree resolution from original CCAM outputs using conservative interpolation.  </li> <li> Period Coverage: <ul> <li>Training Data: 1960-2100 (Depending on Experiment, see Table 1 for list of experiment configurations)</li> <li>Validation Data: 1985-2014 + 2070-2099 (to compute the climate change signal)</li> </ul> </li> <li> Models: <ul> <li>Training on: ACCESS-CM2</li> <li>Validated on: EC-Earth3, NorESM2-MM, CNRM-CM6-1, AWI-MR-1 </li> </ul> </li> </ul> <h3>File Structure</h3> <ul> <li> Training Data: <ul> <li>Target/Ground Truth (Y): <code>target_ACCESS-CM2_hist_ssp370_pr.nc</code></li> <li>Predictor (X): <code>predictor_ACCESS-CM2_hist_ssp370.nc</code></li> </ul> </li> <li> Evaluation Data: All other GCMs can be accessed in one single file, predictor and target variables have the dimensions (time, lat, lon, GCM). <ul> <li>Target/Ground Truth (Y): <code>Other_GCMs_hist_SSP370_target_fields_pr.nc</code></li> <li>Predictor (X): <code>Other_GCMs_hist_SSP370_predictor_fields.nc</code></li> </ul> </li> </ul> <h2>Methodological Insights</h2> <ul> <li> Regional Climate Model, Our Regional Climate Model training data is from the Conformal Cubic Atmospheric Model (CCAM) which is a global non-hydrostatic atmospheric model renowned for its variable-resolution cubic grid. . For more information about CCAM, please see the following <a href="https://agupubs.onlinelibrary.wiley.com/doi/abs/10.1029/2023JD038530">paper</a>. </li> <li> Predictor and Target Variables: Daily-averaged large-scale prognostic variables, including zonal wind, meridional wind, temperature, and specific humidity, are employed as predictors at the 500mb and 850mb pressure levels. These are normalized (see the GitHub repository for the mean and standard deviation fields). Precipitation is taken as is from CCAM and accumulated for each given day. Static predictors are also used in our model, which is stored in a GitHub repository. </li> <li> Training Framework: Our dataset benefits from the "perfect framework" training strategy, which uses CCAM-coarsened predictor variables. For more information about the perfect and imperfect training frameworks, see the following <a title="review" href="https://journals.ametsoc.org/view/journals/aies/3/2/AIES-D-23-0066.1.xml">review</a> </li> </ul> <table> <tbody> <tr> <td> Algorithm </td> <td> Training Data </td> <td> Period </td> </tr> <tr> <td> Deterministic Baseline </td> <td> Historical </td> <td> 1960-2014 (~21,000 days) </td> </tr> <tr> <td> Deterministic Baseline </td> <td> Future (SSP370) </td> <td> 2044-2099 (~21,000 days) </td> </tr> <tr> <td> Deterministic Baseline </td> <td> Historical and Future (SSP370) </td> <td> 1960-2099 (~51,000 days) </td> </tr> <tr> <td> Residual GAN </td> <td> Historical </td> <td> 1960-2014 </td> </tr> <tr> <td> Residual GAN </td> <td> Future (SSP370) </td> <td> 2044-2099 </td> </tr> <tr> <td> Residual GAN </td> <td> Historical and Future (SSP370) </td> <td> 1960-2099 </td> </tr> </tbody> </table> Table 1: The six RCM emulator experiments performed in this study.

opencc-by-4.0Sep 2024View details →

zenodo32/100

Dataset for "MA-MGAN: Mixed Attention Markovian Generative Adversarial Network for Meteorological Downscaling"

Dataset for "MA-MGAN: Mixed Attention Markovian Generative Adversarial Network for Meteorological Downscaling", including the training set and testing set used by the model, as well as the experimental results in the main text and supplementary Information.

opencc-by-4.0May 2024View details →

zenodo32/100

Dataset: Data augmentation experiments with style-based quantum generative adversarial networks on trapped-ion and superconducting-qubit technologies

Dataset for the following paper: <a href="https://arxiv.org/abs/2405.04401">"Data augmentation experiments with style-based quantum generative adversarial networks on trapped-ion and superconducting-qubit technologies", Julien Baglio, arXiv:2405.04401</a> It contains: <ul> <li>one folder named "data_for_all_plots" containing the raw data for the s, t, and y distributions for all the figures of the paper as well as a Jupyter notebook to generate the figures.</li> <li>one file named "variance_calculations_qGAN.txt" containing the data to calculate the errors for the KL divergences.</li> </ul>

opencc-by-4.0May 2024View details →

zenodo32/100

Accelerating Whole-Sample Polarization-Resolved Second Harmonic Generation imaging in Mammary Gland Tissue via Generative Adversarial Networks

Authors: Arash Aghigh, Jysiane Cardot, Melika Saadat Mohammadi, Gaëtan Jargot, Heide Ibrahim, Isabelle Plante, François Légaré Affiliations:     1.    Centre Énergie Matériaux Télécommunications, Institut National de la Recherche Scientifique, Varennes, Québec, Canada.     2.    Centre Armand-Frappier Santé Biotechnologie, Institut National de la Recherche Scientifique, Laval, Québec, Canada. Corresponding Author: Arash Aghigh, arash.aghigh@inrs.ca Description: This dataset accompanies the research on improving whole-sample Polarization-Resolved Second Harmonic Generation (P-SHG) imaging in mammary gland tissue using Enhanced Super-Resolution Generative Adversarial Networks (ESRGAN). The novel approach significantly reduces imaging time while maintaining high image quality and analytical accuracy, demonstrating a reduction in imaging time by more than 95%. This method also minimizes laser-induced photodamage, lowers costs of optical components, and increases the accessibility and applicability of P-SHG imaging in various fields. Keywords: Polarization-Resolved Second Harmonic Generation, P-SHG, Generative Adversarial Networks, GAN, ESRGAN, Mammary Gland Imaging, Super-Resolution, Image Upscaling, Deep Learning, Biomedical Imaging Funding Information:     •    Canada Foundation for Innovation     •    Fonds de recherche du Québec–Nature et technologies     •    Natural Sciences and Engineering Research Council of Canada     •    New Frontiers Research Fund     •    NSERC CREATE program (scholarship for Arash Aghigh) Related Identifiers:     •    GitHub repository for ChaiNNer program: https://github.com/chaiNNer-org/chaiNNer     •    Download links for models used: https://openmodeldb.info Additional Information: Animal studies were conducted according to the procedures provided by the Canadian Council on Animal Care. The protocol (2005-02) was reviewed and approved by the Institutional Committee for Animal Protection of the Laboratoire National de Biologie Expérimentale (LNBE), the animal facilities based at the Institut National de Recherche Scientifique (INRS).

opencc-by-4.0Dec 2023View details →

zenodo32/100

Resolution Enhancement and Deblurring of Porous Media μ-CT Images based on Super Resolution Generative Adversarial Network

The above is the Data2 of Super Resolution Generative Adversarial Network based on High-Resolution Representation Learning.

opencc-by-4.0Aug 2023View details →

ClinicalTrials.gov32/100

Modelling Tau Distribution From DTI With Generative Adversarial Network for Alzheimer's Disease Diagnosis

ClinicalTrials.gov study NCT05020626. IPD Sharing: Not stated. Countries: 1. Publications: 33.

restrictedIPD-UNDECIDEDFeb 2026View details →

dryad32/100

Animal acoustic identification, denoising, and source separation using generative adversarial networks

Open the record for dataset details and reuse information.

publicAug 2025View details →

zenodo28/100

Synthetic histology images of colorectal cancer, generated by conditional generative adversarial networks

These are generated (synthetic) histology images of colorectal cancer. These images were generated by conditional GANs and are in two classes: MSIH (microsatellite instable high) and nonMSIH. There are two sets: one set with 10K images per class and another one with 75K images per class. All images are RGB, 512x512 px at a resolution of 0.5 micrometers per pixel. For more information, please stay tuned for our upcoming manuscript on www.kather.ai.

opencc-by-4.0Jul 2020View details →

ScienceDex guides

Understand access before you commit

These curated guides explain access requirements, typical timelines, costs, and reuse considerations for widely used research datasets.

Compare curated datasets

Allen Brain Atlas

Allen Brain Atlas is an Allen Institute collection of brain map atlases, datasets, APIs, and analysis tools covering mouse, human, and non-human primate brain resources.

allen-brain-atlas

neuroscienceopenDocumentation, web resources, and API references are available online.

Last verified 2026-04-30Open record

Annotated Behaviour and Observability Dataset (ABODe)

ABODe is a University of Edinburgh DataShare dataset for behavior classification in group-housed mice using home-cage video, identities, bounding boxes, ground-plate positions, and annotator labels.

abode-home-cage

behavioral-neuroscienceopenThe DataShare record exposes download links for annotations, documentation, license text, and the zipped per-snippet data directory.

Last verified 2026-04-30Open record

DANDI Archive for NWB datasets

DANDI is a BRAIN Initiative archive for publishing and sharing neurophysiology data, including electrophysiology, optophysiology, and behavioral data packaged as NWB and related standards.

dandi-nwb

electrophysiologyopenPublished Dandiset metadata and archive endpoints are available through the production DANDI API.

Last verified 2026-04-30Open record

International Brain Laboratory public data

The International Brain Laboratory public data releases expose standardized mouse decision-making experiments, including Neuropixels recordings, widefield calcium imaging, behavior, and session metadata accessed through the ONE API.

ibl

behavioral-neuroscienceopenPublic sessions can be searched and loaded from the IBL public data server through ONE.

Last verified 2026-04-29Open record

OpenNeuro

OpenNeuro is a free, open platform for sharing neuroimaging datasets, with public search, dataset pages, and download paths for web, S3, DataLad, and the OpenNeuro CLI.

openneuro

neuroscienceopenPublished datasets are available on demand over the internet.

Last verified 2026-04-29Open record