You can download the pre-trained weights for this model from https://drive.google.com/file/d/1DXvg3UW9V8IznDC280l8zGMdbPCz8Tty/view?usp=sharing
To test:
To test the model you can unzip test_selected.zip and refer the path to it in the InceptionNasnet_test.py file. Download weights and set weight file path also. Run the file using the command
python InceptionNasnet_test.py
To train:
To train the model you need training data which is huge and so not shared but you can follow the lines of creating the train_split_v3.txt for your own training data and train from scratch. Or do transfer learning with the weights provided here. All you need to do is
- create train_split.txt space separated file for your data which contains the labels
- set train images path and test images path in the InceptionNasnet_train.py file.
- run command "python InceptionNasnet_train.py"