Aivia Software
Bioimage.io models in Aivia
The bioimaging community produces and increasing number of deep learning models which they make available publicly for download. The BioImage Model Zoo’s repository currently allows users of community open-source tools to search for, browse and download DL models. In Aivia, we control the versions to be compatible with our software, thus, you can download them here: Deep Learning Model Library | Aivia
These models have been tested via Drag & Drop and can be used in Aivia in the recipe console, producing channels as output.
List of available bioimage models
Information to retrieve the correct model:
Find your compatible version in Aivia Compatible versions, then click in the Link to download it.
Models 001 to 004
Number | Name | Aivia compatible versions | Link to website to download | Model name | Framework | Description |
|---|---|---|---|---|---|---|
001 | Segmentation for bacteria - Tensorflow version | 12.1, 13.0, 14.0 | bioimageio-001-Segmentation-bacteria-105281_zenodo_7261974.aiviadl | tensorflow 2.6.0, keras 2.6.0 | 2D U-Net model segments the contour, foreground and background of Bacillus Subtilis bacteria imaged with Widefield microscopy images. https://bioimage.io/#/?id=10.5281%2Fzenodo.7261974 | |
001 | Segmentation for bacteria - Keras Pytorch version | 15.0 | Pending | bioimageio-001-Segmentation-bacteria-105281_zenodo_7261974_keras.aiviadl | pytorch 2.2.0, keras 3.0.5 | 2D U-Net model segments the contour, foreground and background of Bacillus Subtilis bacteria imaged with Widefield microscopy images. |
002 | Pancreatic cell segmentation - Tensorflow version | 12.1, 13.0, 14.0 | bioimageio-002-Pancreatic-cell_segmentation-105281_zenodo_5914248.aiviadl | tensorflow 2.6.0, keras 2.6.0 | U-Net trained to segment phase contrast microscopy images of pancreatic stem cells on a 2D polystyrene substrate. | |
002 | Pancreatic cell segmentation - Keras Pytorch version | 15.0 | Pending | bioimageio-002-Pancreatic-cell_segmentation-105281_zenodo_5914248_keras.aiviadl | pytorch 2.2.0, keras 3.0.5 | U-Net trained to segment phase contrast microscopy images of pancreatic stem cells on a 2D polystyrene substrate. |
003 | Nuclei cell segmentation | 12.1, 13.0,14.0, 15.0 | bioimageio-003-HPA_nucleus-105281_zenodo_6200999.aiviadl | pytorch 1.11.0+cuda11.3 | Nuclei segmentation model for segmenting images from the Human Protein Atlas | |
004 | Body Cell segmentation | 12.1, 13.0, 14.0, 15.0 | HPA Cell Segmentation - BioImage.io
| bioimageio-004-HPA_cell_body-105281_zenodo_6200635.aiviadl | pytorch 1.11.0+cuda11.3 | Cell Body segmentation model for segmenting images from the Human Protein Atlas |
Number | Name | Aivia compatible versions | Link to website to download | Model name | Framework | Description |
|---|---|---|---|---|---|---|
005 | Neuron Segmentation in EM - CREMI Challenge | 13.0 to 15.0 | Pending | bioimageio-005-CREMI_3D_EM-105281_zenodo_5874741.aiviadl | pytorch 1.11.0+cuda11.3 | EM neuron segmentation for CREMI 3D challenge. |
006 | 3D UNet Mouse Embryo Fixed | 13.0 to 15.0 | Pending | bioimageio-006-3D_MouseEmbryoFixed-105281-zenodo_6383429.aiviadl | pytorch 1.11.0+cuda11.3 | A 3D U-Net trained to predict the nuclei and their boundaries in fixed confocal images of developing mouse embryo. |
007 | 3D Unet Mouse Embryo Live | 13.0 to 15.0 | Pending | bioimageio-007-3DMouseEmbryoSegmentation-105281_zenodo_6384845.aiviadl | pytorch 1.12.1+cuda11.3 | A 3D U-Net trained to predict the cell boundaries in live light sheet images of developing mouse embryo. |
008 | NucleiSegmentationBoundary | 13.0 to 15.0 | Pending | bioimageio-008-NucleiBoundary_Segmentation-105281_zenodo_5764892 | pytorch 1.10.2+cuda11.3 | This model segments nuclei in fluorescence microscopy images. It predicts boundary maps and foreground probabilities for nucleus segmentation in different light microscopy modalities, mainly with DAPI staining. https://bioimage.io/#/?id=10.5281%2Fzenodo.5764892 |
009 | CovidIFCellSegmentationBoundaries | 13.0 to 15.0 | Pending | bioimageio-009-CovidIFCellSegmentationBoundary-105281_zenodo_5847355.aiviadl | pytorch 1.11.0+cuda11.3 | This model segments cells in immunofluorescence microscopy images. It predicts boundary maps and foreground probabilities and was trained on Vero E6 cells imaged with a high-throughput-microscope, as part of a Covid19 antibody test. |
010 | LiveCellSegmentationBoundary | 13.0 to 15.0 | Pending | bioimageio-010-LiveCell_Segmentation-105281_zenodo_5869899.aiviadl | pytorch 1.11.0+cuda11.3 | This model segments cells in phase-contrast microscopy images, which are often used in live-cell imaging. It predicts boundary maps and foreground probabilities. The boundaries can be processed e.g. with Multicut or Watershed to obtain an instance segmentation. |
011 | 2D Unet Arabidopsis Ovules | 13.0 to 15. | Pending | bioimageio-011-2D_ArabidopsisOvule-105281-zenodo-7805067.aiviadl | pytorch 1.11.0+cuda11.3 | A a variant of 2D U-Net trained to predict the cell boundaries in confocal stacks of Arabidopsis ovules. |
012 | 2D UNet Arabidopsis Apical | 13.0 to 15. | Pending | bioimageio-012-2D_ArabidopsisApicalStem-105281-zenodo-6334881.aiviadl | pytorch 1.11.0+cuda11.3 | 2D Unet trained on z-slices of confocal images of Arabidopsis thaliana apical stem cells |
013 | 3D Unet Lateral Root Primordium | 13.0 to 15. | Pending | bioimageio-013-3D_ArabidopsisLateralRoot-105281-zenodo-6334777.aiviadl | pytorch 1.11.0+cuda11.3 | A 3D U-Net trained to predict the cell boundaries in lightsheet stacks of Arabidopsis Lateral Root Primordia. |
014 | 3D Unet Arabidopsis Ovules | 13.0 to 15. | Pending | bioimageio-014-3D_ArabidopsisOvuleNuclei-105281-zenodo-7772662.aiviadl | pytorch 1.11.0+cuda11.3 | Unet trained on confocal images of Arabidopsis Ovules nuclei stain with BCEDiceLoss. The network predicts 1 channel: nuclei probability maps. |
015 | 3D UNet Arabidopsis Ovules | 13.0 to 15. | Pending | bioimageio-015-3D_ArabidopsisOvuleCells-105281-zenodo-6334583.aiviadl | pytorch 1.11.0+cuda11.3 | A 3d U-Net trained to predict the cell boundaries in confocal stacks of Arabidopsis ovules. |
016 | 3D Unet Arabidopsis Apical Stem | 13.0 to 15. | Pending
| bioimageio-016-3D_ArabidopsisApicalStem-105281-zenodo-6346511.aiviadl | pytorch 1.11.0+cuda11.3 | 3D Unet trained on confocal images of Arabidopsis thaliana apical stem cell. |
EM Models
Number | Name | Biozoo model | Link to website | Aivia working versions | Framework | Description |
|---|---|---|---|---|---|---|
017 | MitochondriaEMSegmentation | bioimageio-017-Mitochondria_EM_Boundary-105281_zenodo_5874741.aiviadl | Pending | 15 | pytorch 2.5.1+cu11.8 | Segments Mitochondria in EM neuron images (3D). |
018 | PlatynereisEMnucleiSegmentationBoundaryModel | bioimageio-018-Platynereis_Nuclei_EM_Boundary-105281_zenodo_6028097.aiviadl | Pending | 15 | pytorch 2.5.1+cu11.8 | Segments cell nuclei + boundaries Platynereis 3D |
019 | PlatynereisEMcellSegmentationBoundaryModel | bioimageio-019-Platynereis_Cells_EM_Boundary-105281_zenodo_6028280.aiviadl | Pending | 15 | pytorch 2.5.1+cu11.8 | Segments cell boundaries Platynereis 3D. |
020 | MitochondriaEMSegmentation2D | bioimageio-020-Mitochondria_EM_Boundary-105281_zenodo_6406803.aiviadl | Pending | 15 | pytorch 2.5.1+cu11.8 | Segments mitochondria in serial sections of TEM (ssTEM). |
021 | CebraNET | bioimageio-021-CebraNET-105281_zenodo_7274275.aiviadl | Pending | 15 | pytorch 2.5.1+cu11.8 | Generates membrane probability from SEM 3D cubes. |
022 | MitoNet | bioimageio-022-MitoNet-stupendous-sheep.aiviadl | Pending | 15 | pytorch 2.5.1+cu11.8 | This model generates probability of mitochondria |
For python experts: Aivia Version - AIS Version
Internally, AIS is a wheel that has the framework that run the models. Here is some information:
Aivia Version | AIS version |
|---|---|
Aivia 12.1.039591 | aiviaserving-0.0.5-py2.py3-none-any.whl Python 3.9 |
Aivia 13 | aiviaserving-0.1.30-py2.py3-none-any.whl Aivia 13 was released with aivia-serving-0.1.24, and 13.5 with 0.1.30. Use 0.1.30 to avoid some bugs. Python 3.9 |
Aivia 15 | aiviaserving-0.1.43-py2.py3-none-any.whl, not compatible with tensorflow (keras/pytorch) Python 3.12.3 |
References
Paper by the autors of bioimage.io https://www.biorxiv.org/content/10.1101/2022.06.07.495102v1.full.pdf