Tiskus, Edvinas · Tiškuvienė, Rūta · Bučas, Martynas · et al.
37 files · 1.6 GB · hdf5, tiff, torchdeclared
hybrid · semantic + lexical · 12 datasets ranked · 0.73s
Tiskus, Edvinas · Tiškuvienė, Rūta · Bučas, Martynas · et al.
37 files · 1.6 GB · hdf5, tiff, torchdeclared
This record contains the labeled data, trained models, and analysis code supporting the article "Comparing a Vision Foundation Model (DINOv3) and a Task-Specific U-Net for Mapping Emergent Aquatic Vegetation from Fused UAV Multispectral and LiDAR Data" (Remote Sensing in Ecology and Conservation). Contents: - masks/ : georeferenced ground-truth segmentation masks (five classes: aquatic vegetation, water, sand, other objects, background), aligned to the fused UAV orthomosaics and spanning 13 sites across nine Lithuanian waterbodies surveyed between May and August 2024. - models/ : the two final trained segmentation models, a Keras/HDF5 U-Net and a PyTorch DINOv3 model. - code/ : Python scripts for training, evaluation, the label-efficiency experiment, and full-scene prediction. The fused 9-band orthomosaics (five-band multispectral, RGB, and a LiDAR canopy height model; approximately 62 GB) are archived separately because of their size and are available from the corresponding author on request. The DINOv3 SAT-493M pretrained backbone is distributed by Meta under its own license and is not redistributed here; obtain it from the official DINOv3 release.
Bohl, Michael · Esteban-Medina, Marina · Lenhof, Kerstin · et al.
5 files · 5.0 GB · csv, torch, zipdeclared
This Zenodo record contains all data necessary to reproduce the benchmark results described in the following publication: M. Bohl, M. Esteban-Medina, N. Beerenwinkel, and K. Lenhof, Domain-adaptation deep learning models do not outperform simple baseline models in single-cell anti-cancer drug sensitivity prediction, bioRxiv (2026). Processed bulk and single-cell RNA-Seq datasets with response labels are in processed.zip. scATD model weights are in checkpoint_fold1_epoch_30.pth Full hyperparameter tuning logs/results are in hyperparam_tuning_results.csv A revised version of the source code (without model weights) is in code.zip. If it gets updated in the future, check the latest version at https://github.com/cbg-ethz/SC-Bulk-Domain-Adaptation/
BROCHARD, Pierre
3 files · 141 MB · torchdeclared
YOLOv26 models specialized in text region (TextRegion) and text line (TextLine) segmentation for medieval manuscripts. Sources: The Manicule corpus: https://nakala.fr/collection/10.34847/nkl.e0ef83vx The Alcar-HOME database: https://zenodo.org/record/5600884 The e-NDP corpus: https://zenodo.org/record/7575693 The Himanis project: https://zenodo.org/record/5535306 OCR ground truth for Caroline Miniscule : https://github.com/rescribe/carolineminuscule-groundtruth Ground Truth for ONB-Cod. 3891 : https://zenodo.org/record/7467249 Cremma Medieval : https://zenodo.org/record/7506657 DISTINGUO : https://doi.org/10.34847/NKL.48AD8B8D and synthetic data. HuggingFace Mirror : https://huggingface.co/LaMOP/Yolo-Seg-TextRegion-TextLine-Manuscript
Meekes, Lisa · Tabaro, Francesco · Bexkens, Michiel · et al.
41 files · 8.2 GB · csv, fasta, pdfdeclared
This record contains the Python software for PEPTiGEN, a tool for generating tryptic peptides from prokaryotic gene sequences and their variants, and the associated antimicrobial resistance (AMR) peptide database. The database is provided as an SQL file and a CSV file containing all genes and predicted peptides. The README file contains explanation of the PEPTiGEN tool. The SQL database schema files contains both the database schema of the SQL database used in the PEPTiGEN analysis as the database schema of the AMR peptide datbase.
Yang, Jiazhuo · Cai, Hongyan · Xu, Xinliang · et al.
3 files · 408 MB · tiff, torchdeclared
This record contains the 10 m farmland shelterbelt distribution products for Northeast China in 2020 and 2025, together with the pretrained deep learning model weights used for farmland shelterbelt inference. The dataset was generated from spring Sentinel-2 surface reflectance composites using B4, B8, and NDVI features and a ResNet-50/CA deep learning model. The two GeoTIFF files represent binary farmland shelterbelt maps, where 1 indicates farmland shelterbelt and 0 indicates non-shelterbelt. The pretrained model weights are provided as ForestNet50V1.pth to support reproducible inference and local model adaptation. The dataset covers major agricultural regions of Northeast China, including Heilongjiang, Jilin, Liaoning, and eastern Inner Mongolia. It is intended for regional- and landscape-scale analyses of farmland shelterbelt distribution, spatial continuity, fragmentation, stage-based change, and ecological engineering assessment.
Anonymous Authors
15 files · 2.2 GB · parquet, rar, torchdeclared
Economic losses caused by extreme climate are not confined to the locations where events occur, but can propagate across regions through physical and economic linkages. Yet existing climate-impact assessment methods remain poorly suited to tracing how shocks spread across space and reshape the geography of economic loss. Here we develop a mechanistically informed multi-scale spatiotemporal autoregressive graph neural network model to quantify spatially cascading climate impacts. The model couples scale-specific, physically structured spatiotemporal autoregressive processes through an adaptive gating mechanism, allowing heterogeneous cross-scale interactions to be learned from data. Model estimation is achieved through tailored graph convolutional neural networks that are mathematically equivalent to spatiotemporal autoregressive models, enabling scalability while preserving transparent parameter interpretation. Monte Carlo simulation experiments show that the model accurately recovers true parameters and distinguishes between scale-dependent processes. Applying the framework to extreme precipitation, we find that large-scale upwind-to-downwind cascades driven by atmospheric circulations dominate aggregated economic losses. A one-standard-deviation increase in log extreme precipitation is associated with a 0.19 percentage-point decline in economic growth rate at the large scale, with 62.3% of the loss arising from spatial cascades. These findings highlight the need for transboundary risk governance that incorporates spatial cascading into climate-extremes monitoring and early-warning. Description of the uploaded file Monte Carlo simulation code data_generator_factors.py: Data generation script for multi-scale Monte Carlo simulation experiments. sarnn_model.py:Implementation of the proposed MS-STARGNNs model architecture definition. train.py:Training pipeline script for the MS-STARGNNs model. Data and spatial weights matrices for empirical analysis global_panel_1deg_std.parquet:Standardized Large-scale (1°) datase; global_panel_2km_std.parquet:Standardized Small-scale (2 km) dataset. W_global_2km_knn8.pt:Small-scale spatial weights matrix based on 8-nearest neighbors (KNN8). W_Large-scale:A large-scale spatial weights matrix derived from moisture transport pathways (2005-2021) mapping_1deg_to_2km.parquet: Correspondence file mapping large-scale (1°) grid cells to small-scale (2km) grid cells. ipcc_region_mapping_coarse_8regions.parquet: Mapping file linking Large-scale grid cells to the 8 IPCC AR6 reference regions. ipcc_region_mapping_fine_8regions.parquet: Mapping file linking Small-scale grid cells to the 8 IPCC AR6 reference regions. Code model.py: Core architecture definitions for empirical analysis. Provides the base classes and computational layers engineered to handle real-world geospatial complexities. All subsequent training scripts import modules from this file. STARGNNs.py: Implementation of the single-scale baseline. Serves as a reference point for evaluating the efficacy of cross-scale feature fusion. MS-STARGNNs_fixed.py: Configuration script for the MS-STARGNNs model utilizing fixed autoregressive coefficients. MS-STARGNNs.py: Configuration script for the MS-STARGNNs model utilizing annually varying autoregressive coefficients. MS-STARGNNs_8 regions.py: Executes the MS-STARGNNs model with decoupled regional parameters, loading unique autoregressive weights and βvectors for each IPCC region. Implements null value handling for regions lacking observational data (e.g., Antarctica).
zhang, Wenjun
10 files · 459 MB · torchdeclared
This archive contains full training dataset, preprocessing scaler .pkl files and five groups of pre-trained neural network checkpoints (.pth) for reproducing all results in the manuscript. Corresponding source code repository on GitHub: https://github.com/wenjunzhang2020/Prediction-of-combined-effects-of-binary-mixed-systems .
Mutluel, Abdullah Mevlüt
12 files · 449 KB · pdf, torchdeclared
Code and data accompanying the manuscript "Disentangling Electro-osmotic Drag and Back-Diffusion Water Fluxes in Polymer Electrolyte Membranes: A Physics-Informed Neural Network for Net-Flux Inversion" submitted to the Journal of the Electrochemical Society. Contents: - gen_data_lit.py: synthetic benchmark generation. Solves the through-plane membrane water transport boundary-value problem for 24 operating conditions using the Springer drag coefficient and the Nguyen-White diffusion correlation. - pinn_lit.py: physics-informed neural network training. Recovers n_d(lambda) and D_w(lambda) from net-flux data with a hard conservation constraint and a single anchor point. - plots_lit.py: reproduces all manuscript figures (Figs. 1-3). - rev_lib.py: validation experiments, including trend-line baseline comparison, noise robustness, non-monotonic diffusion coefficient recovery, anchor ablation, multi-seed statistics, and held-out condition tests (Fig. 4 and Table 2). - ga_final.py: graphical abstract. - data_lit.json: generated benchmark data (24 operating conditions with internal water-content profiles and net fluxes). - model_lit.pt: trained PyTorch model weights. - manuscript.tex and figure PDFs. Requirements: Python 3 with PyTorch, NumPy, SciPy, and Matplotlib. Run gen_data_lit.py first, then pinn_lit.py, and then plots_lit.py to reproduce the results from scratch, or load model_lit.pt directly with the Model class defined in pinn_lit.py.
S.S, Ashwin · Minami, Katsuhiko · Nakazato, Kako
11 files · 124 MB · jpeg, torch, zipdeclared
Supplementary code for: Katsuhiko Minami, Kako Nakazato, Sachiko Tamura, S. S. Ashwin, Kazuhiro Maeshima* Machine learning-assisted Repli-Histo labeling reveals distinct transcription-dependent constraints on chromatin motion in living cells (2026).
Gupte, Nihar · Miller, M. Coleman · Udall, Rhiannon · et al.
36 files · 36 GB · hdf5, torchdeclared
Data release for the DINGO O4a eccentricity paper. It contains the per-event parameter-estimation products, population selection function, and hierarchical-inference posteriors needed to reproduce every figure, table, and number in the paper, plus the trained DINGO neural networks used for the analyses. Event data : eccentric, quasicircular, and precessing per-event posterior samples (posteriors_eccentric.h5, posteriors_quasicircular.h5, posteriors_precessing.h5); slimmed log-uniform-eccentricity-prior posteriors used as the hierarchical-likelihood input (posteriors_log_uniform_eccentric.h5); per-event posteriors reweighted by the population-informed posterior (posteriors_population_reweighted.h5); per-event summary statistics with pre-computed Bayes factors (summary_statistics.h5); e_gw conversions (egw_conversions.h5); and the eccentricity-mean-anomaly prior hull (e_zeta_prior_hull.h5). Selection function : the injection p_draw dataframe with detection probabilities including the analysis-window factor (injection_p_draw.h5), a fixed-injection eccentricity sweep (fixed_injection_ecc_sweep.h5), and matched-filter survival-function data (survival_function.h5). Hierarchical inference : the selection-corrected velocity-dispersion posterior marginalized over the GWTC-4 mass/spin/redshift hyperposterior (sigma_posterior.h5), the capture-eccentricity lookup table (capture_ecc_table.h5), the external GWTC-4 hyperposterior fit (gwtc4_hyperposterior.h5), and the GC/NSC branching-fraction posterior (branching_fraction_posterior.h5). Glitch analyses : glitch-marginalized posteriors for GW190701, GW231114_043211, and GW231223_032836. Networks : trained DINGO networks (SEOBNRv5EHM, SEOBNRv5HM, SEOBNRv5PHM) with their training settings; see MODEL_MANIFEST.md. Zenodo stores files flat; the companion code maps each file into the foldered layout the notebooks expect. Code to download the data and reproduce all figures: github.com/nihargupte-ph/o4a-eccentricity , archived at doi:10.5281/zenodo.21221948 .
Liu, Licheng
2 files · 5.2 MB · torchdeclared
Khalid, Adnan
1 files · 129 KB · torchdeclared