Image Segmentation
Transformers
English
Inference Endpoints
File size: 1,279 Bytes
7b650b1
 
e8e6847
 
652ec03
2f91a5a
 
652ec03
2f91a5a
 
3d0df08
 
2f91a5a
 
 
67f531b
2f91a5a
652ec03
2f91a5a
652ec03
 
 
 
 
 
 
 
 
 
 
2f91a5a
652ec03
67f531b
 
 
 
 
 
652ec03
 
596974f
652ec03
67f531b
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
---
license: gpl-3.0
language:
- en
pipeline_tag: image-segmentation
---

# 3DL_NuCount model


__Model author__: Fabrice Daian


#### Model description

__3DL_NuCount__ model has been designed by fine tuning a pretrained Stardist3D model [1,2] using a home made dataset [3] in order to assess the number of cells present in a given 3D image stack acquired using an optical microscope. Training and Inference Notebooks are hosted on our Github repo [4].

#### Stardist Training parameters

- patch size:  (48,96,96)
- batch size:  32
- epochs    :  100
- data augmentation : flip/rotation/intensity
- image normalization: normalize channel independantly
- anisotropy:  empirical
- rays      :  96

#### Training dataset parameters
- tile size : (4,63,128,128)
- split     : Train 0.8 / Val 0.2

#### Inference
- patch size     : (784,784,:)
- image size     : (2048,2048,:)
- model name     : __weights_best_1.h5__
- config file    : __config.json__
- threshold file : __thresholds.json__


#### References
- [1] Stardist Project: [Github](https://github.com/stardist/stardist)
- [2] Stardist Paper :  [ArXiv](https://arxiv.org/abs/1908.03636)
- [3] NuCount Training Dataset  : [Zenodo](https://)
- [4] NuCount Github Project: [Github](https://github.com/andysaurin/3DL_NuCount/)