Skip to content

Latest commit

 

History

History
65 lines (48 loc) · 1.9 KB

README_old.md

File metadata and controls

65 lines (48 loc) · 1.9 KB
Raw

U-NET in Pytorch for Image Segmentation

This repo is an implementation of U-Net for penguin colony detection. It is under active development.

This code is written by Hieu-Le.

Note: The current software works well with PyTorch 0.4.

Prerequisites

  • Linux
  • Python 3
  • CPU or NVIDIA GPU + CUDA CuDNN

Getting Started

Installation

  • Install PyTorch 0.4 and dependencies from http://pytorch.org
  • Install Torch vision from the source.
git clone https://github.com/pytorch/vision
cd vision
python setup.py install
pip install visdom
pip install dominate
  • Alternatively, all dependencies can be installed by
pip install -r requirements.txt
  • Clone this repo:
git clone https://github.com/iceberg-project/Penguins/

Prediction

  • Download a pre-trained model at:

https://drive.google.com/file/d/149j5rlynkO1jQTLOMpL5lextHY0ozw6N/view?usp=sharing

Please put the model file to: <checkpoints_dir>/<model_name>/

The one provided here is at the epoch 300 of the model named "v3weakly_unetr_bs96_main_model_ignore_bad"

  • The script to run the testing for a single PNG image:

python predict.py [--params ...]

params:

  • --name: name of the model used for testing
  • --gpu_ids: the gpu used for testing
  • --checkpoints_dir: path to the folder containing the trained models
  • --epoch: which epoch we use to test the model
  • --input_im: path to the input image
  • --output: directory to save the outputs

Example script:

python predict.py --gpu-ids 0 --name v3weakly_unetr_bs96_main_model_ignore_bad --epoch 300 --checkpoints_dir '../checkpoints_CVPR19W/' --output test --testset GE --input_im ../data/Penguins/Test/A/GE01_20120308222215_1050410000422100_12MAR08222215-M1BS-054072905140_01_P002_u08rf3031.png