Input :
- data_path can be data for the training or the testing.
- is_train is a boolean. True for training. False for testing.
- checkpoint_dir
- result_path
- save_parameters is a boolean. Set to 1 to save filters, bias, feature maps.
- config configuration of network.
Output :
- configuration.mat with data_path, result_path, config, is_train
- parameters.mat (if save_parameters = 1)
- RMSE_tab save the evolution of RMSE at each step
- _down : input depth
- _rgb : input intensity
- _up : GT depth
- _sr : output of algorithm
python main.py --data_path = '' --is_train = '' --config= ''--checkpoint_dir= '' --result_path='' --save_parameters = ''
https://li-chongyi.github.io/proj_SR.html
python main.py --data_path='/Users/aliceruget/Documents/PhD/Dataset/Paper_Test_dataset/16' --is_train='0' --config='/Users/aliceruget/Documents/PhD/DepthSR_Net_AR/Configs/cfg_original_scale16.yaml' --checkpoint_dir='/Users/aliceruget/Documents/PhD/DepthSR_Net_AR/Checkpoint/16' --result_path='/Users/aliceruget/Documents/PhD/DepthSR_Net_AR/Results/Paper_Test_dataset_16' --save_parameters='1'