axis=(0, 2, 3))
print mean
print std
train_dataset = supervised_dataset.SupervisedDataset(X_train, y_train)
val_dataset = supervised_dataset.SupervisedDataset(X_val, y_val)
train_iterator = train_dataset.iterator(mode='random_uniform',
batch_size=64,
num_batches=31000)
val_iterator = val_dataset.iterator(mode='random_uniform',
batch_size=64,
num_batches=31000)
# Create object to local contrast normalize a batch.
# Note: Every batch must be normalized before use.
normer = util.Normer3(filter_size=5, num_channels=1)
module_list = [normer]
preprocessor = util.Preprocessor(module_list)
print('Training Model')
for x_batch, y_batch in train_iterator:
#x_batch = preprocessor.run(x_batch)
x_batch = (x_batch - mean) / std
# loop over batch
for i in range(len(x_batch)):
# hide patch for an image
x_batch[i] = hide_patch(x_batch[i])
monitor.start()
log_prob, accuracy = model.train(x_batch, y_batch)
checkpoint_file = args.checkpoint_file
fold = int(args.split)
dataset_path = os.path.join(data_paths.tfd_data_path, 'npy_files/TFD_96/split_'+str(fold))
print 'Checkpoint: %s' % checkpoint_file
print 'Testing on split %d\n' % fold
# Load model
model = SupervisedModel('evaluation', './')
# Load dataset
supervised_data_loader = SupervisedDataLoader(dataset_path)
test_data_container = supervised_data_loader.load(2)
test_data_container.X = numpy.float32(test_data_container.X)
test_data_container.X /= 255.0
test_data_container.X *= 2.0
# Construct evaluator
preprocessor = [util.Normer3(filter_size=5, num_channels=1)]
evaluator = util.Evaluator(model, test_data_container,
checkpoint_file, preprocessor)
# For the inputted checkpoint, compute the overall test accuracy
accuracies = []
print 'Checkpoint: %s' % os.path.split(checkpoint_file)[1]
evaluator.set_checkpoint(checkpoint_file)
accuracy = evaluator.run()
print 'Accuracy: %f\n' % accuracy
accuracies.append(accuracy)
