MFDRNet

Deep learning-based image super-resolution restoration for mobile infrared imaging system

Requirements

python 3.6 pytorch==0.4.0 pillow h5py matplotlib numpy

Usage

First, download the training dataset and place it in IR_ In the data folder;

Second, train the network python train_MFDRN_2.py;

Third, do testing python test_MFDRN_2.py --test_hr_folder ./Test_Datasets/HR/ --test_lr_folder ./Test_Datasets/LR_2/ --output_folder ./results/MFDRN_duibi_x2 --checkpoint ./checkpoint_MFDRN_duibi_x2/epoch_100.pth --upscale 2.