Cdeep3M Pre-trained models

Imaging Type

Cell structure

(54542)

Trained model: • CompImBio Teste
Description: • retrain of 'SBEM and TEM Membranes (Aug-110)' from Matthias Haberl adding imag from brainiac 2015 v2
X Voxelsize: • 5 µm
Y Voxelsize: • 5 µm
Z Voxelsize: •
Cellular component: •
Author
DOI: • https://doi.org/10.7295/W9CDEEP3M54542
Version number: • 1

Download

(50829)

Trained model: • CompImBio Teste1
Description: • retrain of 'SBEM and TEM Membranes (Aug-110)' from Matthias Haberl adding imag from brainiac 2015
X Voxelsize: • 5 µm
Y Voxelsize: • 5 µm
Z Voxelsize: •
Cellular component: •
Author
DOI: • https://doi.org/10.7295/W9CDEEP3M50829
Version number: • 0

Download

(50821)

Trained model: • Synaptic Vesicles SBEM (Aug-12)
Description: • Electron microscopy synaptic vesicles model trained with CDeep3M2 /// Secondary Augmentations: -1 /// Tertiary Augmentations: 2 /// Data: Serial block-face scanning electron microscopy (SBEM) data /// Please cite: Haberl et al., CDeep3M-Plug-and-Play cloud-based deep learning for image segmentation. Nat Methods. 2018 Sep;15(9):677-680. Epub 2018 Aug 31. PMID: 30171236 DOI: 10.1038/s41592-018-0106-z /// and /// Haberl et al., CDeep3M-Preview: Online segmentation using the deep neural network model zoo. DOI: 10.1101/2020.03.26.010660
X Voxelsize: • 4 nm
Y Voxelsize: • 4 nm
Z Voxelsize: • 20 nm
Cellular component: •
Author: • Matthias Haberl
DOI: • https://doi.org/10.7295/W9CDEEP3M50821
Version number: • 2.1.5

Download

SBEM, TEM Membranes (50789)

Trained model: • SBEM and TEM Membranes (Aug-110)
Description: • Electron microscopy membrane model trained 60k iterations with CDeep3M2 /// Secondary Augmentations: -1 /// Tertiary Augmentations: 10 /// Data: Transmission electron microscopy and serial block-face scanning electron microscopy data /// Please cite: Haberl et al., CDeep3M-Plug-and-Play cloud-based deep learning for image segmentation. Nat Methods. 2018 Sep;15(9):677-680. Epub 2018 Aug 31. PMID: 30171236 DOI: 10.1038/s41592-018-0106-z /// and /// Haberl et al., CDeep3M-Preview: Online segmentation using the deep neural network model zoo. DOI: 10.1101/2020.03.26.010660
X Voxelsize: • 5 nm
Y Voxelsize: • 5 nm
Z Voxelsize: • 50 nm
Cellular component: • Membranes
Author: • Matthias Haberl
DOI: • https://doi.org/10.7295/W9CDEEP3M50789
Version number: • 2.1.1

Download

SBEM, TEM Membranes (Aug05) (50772)

Trained model: • SBEM, TEM Membranes (Aug05)
Description: • Electron microscopy membrane model trained with CDeep3M2 /// Secondary Augmentations: 0 /// Tertiary Augmentations: 5 /// Data: Transmission electron microscopy and serial block-face scanning electron microscopy data /// Please cite: Haberl et al., CDeep3M-Plug-and-Play cloud-based deep learning for image segmentation. Nat Methods. 2018 Sep;15(9):677-680. Epub 2018 Aug 31. PMID: 30171236 DOI: 10.1038/s41592-018-0106-z /// and /// Haberl et al., CDeep3M-Preview: Online segmentation using the deep neural network model zoo. DOI: 10.1101/2020.03.26.010660
X Voxelsize: • 5 nm
Y Voxelsize: • 5 nm
Z Voxelsize: • 50 nm
Cellular component: • SBEM, TEM Membranes (Aug05)
Author: • Matthias Haberl
DOI: • https://doi.org/10.7295/W9CDEEP3M50772
Version number: • 2.1.0

Download

SBEM, TEM Membranes (50749)

Trained model: • SBEM and TEM Membranes (Aug-10)
Description: • Electron microscopy membrane model trained with CDeep3M2 /// Secondary Augmentations: -1 /// Tertiary Augmentations: 0 /// Data: Transmission electron microscopy and serial block-face scanning electron microscopy data /// Please cite: Haberl et al., CDeep3M-Plug-and-Play cloud-based deep learning for image segmentation. Nat Methods. 2018 Sep;15(9):677-680. Epub 2018 Aug 31. PMID: 30171236 DOI: 10.1038/s41592-018-0106-z /// and /// Haberl et al., CDeep3M-Preview: Online segmentation using the deep neural network model zoo. DOI: 10.1101/2020.03.26.010660
X Voxelsize: • 5 nm
Y Voxelsize: • 5 nm
Z Voxelsize: • 50 µm
Cellular component: • Membranes
Author: • Matthias Haberl
DOI: • https://doi.org/10.7295/W9CDEEP3M50749
Version number: • 2.1.0

Download

SBEM Synapses (denoised) (50725)

Trained model: • SBEM Synapses (denoised)
Description: • Model trained with multiple denoised SBEM datasets, to generally label the structural features of synapses, including postsynaptic densities and synaptic vesicles docking at the presynaptic site.
X Voxelsize: • 12 nm
Y Voxelsize: • 12 nm
Z Voxelsize: • 60 nm
Cellular component: • SBEM Synapses (denoised)
Author: • Matthias Haberl; • Matthew Madany
DOI: • https://doi.org/10.7295/W9CDEEP3M50725
Version number: • 1.6.3rc2

Download

SBEM Membranes (50692)

Trained model: • SBEM Membranes (denoised, resize augm.)
Description: • Electron microscopy membrane model trained with CDeep3M1.6 until 50k iterations Transfer learning applied from 50k-70k iterations with CDeep3M2 Secondary Augmentations: -1 Secondary Augmentations: 5 Data: Transmission electron microscopy and serial block-face scanning electron microscopy data. Please cite: Haberl et al., CDeep3M-Plug-and-Play cloud-based deep learning for image segmentation. Nat Methods. 2018 Sep;15(9):677-680. doi: 10.1038/s41592-018-0106-z. Epub 2018 Aug 31. PMID: 30171236 DOI: 10.1038/s41592-018-0106-z and Haberl et al., CDeep3M-Preview: Online segmentation using the deep neural network model zoo. DOI: 10.1101/2020.03.26.010660
X Voxelsize: • 5 nm
Y Voxelsize: • 5 nm
Z Voxelsize: • 70 nm
Cellular component: • Membranes
Author: • Matthias Haberl
DOI: • https://doi.org/10.7295/W9CDEEP3M50692
Version number: • 1.6.3rc2

Download

Mitochondria (50689)

Trained model: • SBEM Mitochondria (denoised)
Description: • A broadly trained model for mitochondria segmentation based on SBEM and TEM datasets /// Please cite: Haberl et al., CDeep3M-Plug-and-Play cloud-based deep learning for image segmentation. Nat Methods. 2018 Sep;15(9):677-680. Epub 2018 Aug 31. PMID: 30171236 DOI: 10.1038/s41592-018-0106-z
X Voxelsize: • 4.8 nm
Y Voxelsize: • 4.8 nm
Z Voxelsize: • 70 nm
Cellular component: • Mitochondria
Author: • Matthias Haberl
DOI: • https://doi.org/10.7295/W9CDEEP3M50689
Version number: • 1.6.3rc2

Download

SBEM Membranes (50687)

Trained model: • SBEM Membranes (denoised)
Description: • SBEM Membranes training 50k-55k with denoised images Cite: Haberl et al., CDeep3M-Plug-and-Play cloud-based deep learning for image segmentation. Nat Methods. 2018 Sep;15(9):677-680. Epub 2018 Aug 31. PMID: 30171236. DOI: 10.1038/s41592-018-0106-z.
X Voxelsize: • 4.8 nm
Y Voxelsize: • 4.8 nm
Z Voxelsize: • 70 nm
Cellular component: • Membranes
Author: • Matthias Haberl
DOI: • https://doi.org/10.7295/W9CDEEP3M50687
Version number: • 1.6.3rc2

Download

SBEM (50685)

Trained model: • Synapses
Description: • Model trained to generally label the structural features of a synapse such as synaptic vesicles docking at the membrane and post-synaptic density.
X Voxelsize: • 12 nm
Y Voxelsize: • 12 nm
Z Voxelsize: • 40 nm
Cellular component: •
Author: • Matthew Madany
DOI: • https://doi.org/10.7295/W9CDEEP3M50685
Version number: • 1.6.2

Download

Serial section electron tomography Membranes (50683)

Trained model: • Tomography Membranes
Description: • Trained with CDeep3M 1.6.3 (used in doi: 10.1038/s41592-018-0106-z) Sample: High-Pressure Frozen, freeze substituted tissue Data: Serial section multi-tilt electron tomography Target: Membranes
X Voxelsize: • 1.6 nm
Y Voxelsize: • 1.6 nm
Z Voxelsize: • 1.6 nm
Cellular component: • Membranes
Author: • Matthias G Haberl
DOI: • https://doi.org/10.7295/W9CDEEP3M50683
Version number: • 1.2.0

Download

SBEM Synaptic Vesicles (50682)

Trained model: • SBEM synaptic vesicles
Description: • A trained model for segmentation of synaptic vesicles based on SBEM datasets Please cite: Haberl et al., CDeep3M-Plug-and-Play cloud-based deep learning for image segmentation. Nat Methods. 2018 Sep;15(9):677-680. doi: 10.1038/s41592-018-0106-z. Epub 2018 Aug 31. PMID: 30171236 DOI: 10.1038/s41592-018-0106-z
X Voxelsize: • 2.4 nm
Y Voxelsize: • 2.4 nm
Z Voxelsize: • 24 nm
Cellular component: • Synaptic Vesicles
Author: • Matthias G Haberl
DOI: • https://doi.org/10.7295/W9CDEEP3M50682
Version number: • 1.2.0

Download

SBEM mitochondria (50681)

Trained model: • SBEM and ssTEM Mitochondria
Description: • A broadly trained model for mitochondria segmentation Trained with CDeep3M 1.6.3 (used in doi: 10.1038/s41592-018-0106-z) Data: SBEM and TEM Please cite: Haberl et al., CDeep3M-Plug-and-Play cloud-based deep learning for image segmentation. Nat Methods. 2018 Sep;15(9):677-680. doi: 10.1038/s41592-018-0106-z. Epub 2018 Aug 31. PMID: 30171236 DOI: 10.1038/s41592-018-0106-z
X Voxelsize: • 4.8 nm
Y Voxelsize: • 4.8 nm
Z Voxelsize: • 70 nm
Cellular component: • mitochondria
Author: • Matt Haberl
DOI: • https://doi.org/10.7295/W9CDEEP3M50681
Version number: • 1.2.0

Download

TEM (50673)

Trained model: • SBEM and TEM membranes
Description: • Transmission electron microscopy, serial block-face scanning electron microscopy Trained with CDeep3M 1.6.3 (used in doi: 10.1038/s41592-018-0106-z) Please cite: Haberl et al., CDeep3M-Plug-and-Play cloud-based deep learning for image segmentation. Nat Methods. 2018 Sep;15(9):677-680. doi: 10.1038/s41592-018-0106-z. Epub 2018 Aug 31. PMID: 30171236 DOI: 10.1038/s41592-018-0106-z
X Voxelsize: • 5 nm
Y Voxelsize: • 5 nm
Z Voxelsize: •
Cellular component: •
Author: • Matt Haberl
DOI: • https://doi.org/10.7295/W9CDEEP3M50673
Version number: • 1.2.0rc1

Download