search

Jakaria08 / EESRGAN

Small-Object Detection in Remote Sensing (satellite) Images with End-to-End Edge-Enhanced GAN and Object Detector Network
GNU General Public License v3.0
279 stars 70 forks source link

ValueError: num_samples should be a positive integer value, but got num_samples=0 #11

Closed hopesala closed 4 years ago

hopesala commented 4 years ago

I download the file DetectionPatches_256x256,and then generate the HR / LR directory using file scripts_GAN_HR-LR.py and the HR / LR directory are generated as expected, however when I run the python train.py -c config_GAN.json command after modify the config_GAN.json"data_dir_GT": "/media/cao/WINDOWS/Users/123/Downloads/DetectionPatches_256x256/Potsdam_ISPRS/HR/x4/", "data_dir_LQ": "/media/cao/WINDOWS/Users/123/Downloads/DetectionPatches_256x256/Potsdam_ISPRS/LR/x4/", "FRCNN_model": "/media/cao/WINDOWS/Users/123/Downloads/EESRGAN-master/saved/FRCNN_model_LR_LR_cowc/", "pretrain_model_G": "/media/cao/WINDOWS/Users/123/Downloads/EESRGAN-master/saved/pretrained_models_EESRGAN_FRCNN/170000_G.pth", "pretrain_model_D": "/media/cao/WINDOWS/Users/123/Downloads/EESRGAN-master/saved/pretrained_models_EESRGAN_FRCNN/170000_D.pth", "pretrain_model_FRCNN": "/media/cao/WINDOWS/Users/123/Downloads/EESRGAN-master/saved/pretrained_models_EESRGAN_FRCNN/170000_FRCNN.pth", "pretrain_model_FRCNN_LR_LR": "/media/cao/WINDOWS/Users/123/Downloads/EESRGAN-master/saved/FRCNN_model_LR_LR_cowc/0_FRCNN_LR_LR.pth",

I still got the error :ValueError: num_samples should be a positive integer value, but got num_samples=0

Is there somewhere I forget to do or something is wrong? my Chinese English(Chinalish) is poor, thanks you very much

Jakaria08 commented 4 years ago

It seems like the program is not getting the correct directory. Make sure to include the complete path of the directory. You can print the path here to make sure your path is correct: https://github.com/Jakaria08/EESRGAN/blob/c646296f68c1158c62e6bf9eca5fc82103e4f120/scripts_for_datasets/COWC_EESRGAN_FRCNN_dataset.py#L18

hopesala commented 4 years ago

thanks, the problem is solved, it is the path problem, I passed in the wrong path in the train.py: valid_data_loader = module_data.COWCGANFrcnnDataLoader

shazaizai commented 10 months ago

"pretrain_model_FRCNN_LR_LR": "/media/cao/WINDOWS/Users/123/Downloads/EESRGAN-master/saved/FRCNN_model_LR_LR_cowc/0_FRCNN_LR_LR.pth",where should i download？