busrasi/nativelibwithqt

1.Introduction

This document provides a procedure to train an CDCN model to detect liveness from images.

2. Preparing the enviroment

2.1.Get and install CDCN

Create a main folder for project:

mkdir CDCN-Model 

Download:

cd CDCN-Model 
git clone https://github.com/ZitongYu/CDCN.git

2.2 Create a folder for Python virtual environments

2.2.1.Download python3.8 -venv to create environment:

You must be in ../CDCN-Model folder.

sudo apt install python3.8-venv

2.2.2.Create python environment

python3 -m venv cdcn-env

2.2.3.Activate python environment

source cdcn-env/bin/activate

3.Install essential packages that will be used by CDCN building

Install required packages:

pip3 --no-cache-dir install torchvision
pip install matplotlib
pip3 --no-cache-dir install pandas
pip install scikit-build
pip install opencv-python
pip install imgaug
pip --no-cache-dir install sklearn

You can install all required packages with following command.

pip install -r requirements.txt

4. Building

4.1. Fix Bugs

• First of all, this repository have an indentation error in line 242 of /CDCN/CVPR2020_paper_codes/ train_CDCN.py file. We must fix that problem.

• Secondly, line 246 is "model = model.cuda()" in train_CDCN.py file and we changed to be comment line this line and add the following are to force CPU to be used :

    device=torch.device("cpu")
    	
    model.to(device)
  

• Finally, add following codes to line 203

    os.environ["CUDA_VISIBLE_DEVICES"]=""
  

4.2. Building Training File

cd CDCN/CVPR2020_paper_codes
python3 train_CDCN.py

5. Create Map Images

In this step, we realize that we need map_images, so we will create mapped images from our dataset. We will use repository of PRNet-Depth-Generation to generate mapped images.

5.1. Preparing the enviroment

5.1.1 Get and install PRNet-Depth-Generation

Create a main folder for project:

mkdir PRNet-Map-Images-Generation 

Download:

cd PRNet-Map-Images-Generation 
git clone https://github.com/clks-wzz/PRNet-Depth-Generation.git

5.1.2 Create a folder for Python virtual environments

5.1.2.1. Download python3.8 -venv to create environment:

You must be in ./PRNet-Map-Images-Generation folder.

sudo apt install python3.8-venv

5.1.2.2. Create python environment

python3 -m venv prnet-env

5.1.2.3 Activate python environment

source prnet-env/bin/activate

5.2. Install essential packages that will be used by PRNet-Depth-Generation building

Install required packages:

pip install numpy
pip install scikit-image
pip install scipy
pip install opencv-python

You can install all required packages with following command.

pip install -r requirements.txt

5.3. Building

cd PRNet-Map-Images-Generation/PRNet-Depth-Generation
python3 Generate_Depth_Image.py

NOTE: In this, spot we have some errors about building.