1. Pretrained models
Pretrained models are available in './TSFuse_final.model'
2. Test datasets
- Run
python test_image.py
- Eval
python evaluator.py
1. Data Preparation
Download the MSRS dataset from this link
2. Pre-Processing
- Modify the dataset path and run
python dataprocessing.py
and the processed training dataset is named 'MSRS_train_imgsize_128_stride_200.h5'.
3. Training
- Run
python train_h5py.py
and the trained model is available in './models/'.
or if you do not want to use h5py for training, you can run
python train.py