Sign inSign up

kidzik/he_cell_seg_gpu

By kidzik

•Updated over 5 years ago

Deployment of H&E segmentation by Korsuk Sirinukunwattana. GPU version (5x - 10x faster than CPU)

Image
0

10K+

kidzik/he_cell_seg_gpu repository overview

⁠Getting started

  1. Install docker client
  2. Create ./input, ./output and ./output_tumour directories.
  3. Put your .tiff files in ./input directory.
  4. Run
docker run -v $(pwd)/input:/workspace/HE -v $(pwd)/output:/workspace/Result  -v $(pwd)/output_tumour:/workspace/Result_tumour -it kidzik/he_cell_seg
  1. Find segmentation masks in ./output

For the GPU version, install NVIDIA docker and in point 4. run

docker run --gpus all -v $(pwd)/input:/workspace/HE -v $(pwd)/output:/workspace/Result  -v $(pwd)/output_tumour:/workspace/Result_tumour -it kidzik/he_cell_seg_gpu

Tag summary

Content type

Image

Digest

Size

1.9 GB

Last updated

over 5 years ago

docker pull kidzik/he_cell_seg_gpu