segmentation resources

Github resources

4. Seg-net Cambridge University
Use 16 layers of fully convolutional layers with batch normalizations for semantic segmentation based on caffe framework.

5. V net

6. cascaded FCN

7. Deeplab

8. PSPNet
9. E net
10. Microscopy image browser
12. recognet

Deepmask (Facebook)
13. MS-DNet,  UC Berkeley
14. Mask-RCNN

15. facebook AI research Detectron

16. detectorch


- U-Net: Convolutional Networks for Biomedical Image Segmentation + + + + + + +!topic/caffe-users/cN7ZWepXfkc

Deploy caffemodel for segmentation
How to deploy

caffe train --solver=$solver_file 2>&1 | tee log.txt
using this code will automatically save the log.txt for plotting loss

youtube collections

How to assemble image pair for segmentation for training using lmdb and hdf5

1. lmdb:DIGITS is easy
but limited to 8 bit or RGB png or jpg files

2. hdf5:

lmdb on segnet iffuse

Calculation of dice coefficient and Jaccard index

data curation for coco like annotation


1. check if cuda is available
torch.cuda.is_available() should return true if cuda is installed.