Levigty/ResNet50-Classification

Can be used on your own dataset

ResNet50-Classification

Can be used on your own dataset

Step 1：Your own dataset

You should put images under the data directory:

-dataset\
    -model\
    -train\
          -1\
          -2\
          ...
    -test\
          -1\
          -2\
          ...

If you have already put your images well, then look Step 2. If not, look the following optional or find your own method to put your images like the above.

(optional) or you can use data2img.m to convert imagedata to dataset folder For example you should let your your data be in validation.mat, where the structure of validation is:

-validation
    -data
        -data{i,1} is image data 224×224×3
        -data{i,2} is image label id
    -train: training samples' id
    -test:  testing samples' id

Then run data2image.m to create your own dataset

Step 2: Training

   run train_resnet.py to train the model

Step 3: Testing

   run test_resnet.py to test the model