IDP2021-HSI-Segmentation-by-HRNet-and-ResNet

Coursework project: Industrial Project - UEF, Finland

• Project owners: Thong Nguyen, Borhan Sumon, Agha Danish, Gemal
• This project aims to implement segmentations algorithm for Medical Hyperspectal Imaging (HSI), then implement the GUI for visualization. The DL methods used are HRNet and ResNet.

Dataset

• Place your dataset as follow:

dataset/annotations/
dataset/images/
dataset/masks/

Dataset

• You can run with our model by placing it here:

libs/model/

HRNet

Prerequisites

1. Install requirements pip install -r requirement.txt
2. pip install tensorflow-gpu==2.4.1 (For running with GPU)
3. Change the model configuration depends on your needs in libs/config.py

Training

python tools/train.py train

Evaluation

python tools/evaluation.py 

Open the GUI

python ui/main_ui.py 

ResNet

First, you should go to the folder ResNet and follow the instruction below

Prerequisites

1. Install requirements pip install -r resnet/requirement.txt

Training

python resnet/training.py

Evaluation

python resnet/evaluate.py