def test_MODEL_predictionAugmentated_2D(self): data_aug = Data_Augmentation() pp = Preprocessor(self.data_io2D, batch_size=2, data_aug=data_aug, analysis="fullimage") nn = Neural_Network(preprocessor=pp) for sample in self.sample_list2D: predictions = nn.predict_augmentated(sample) self.assertEqual(len(predictions), 2) for pred in predictions: self.assertEqual(pred.shape, (16, 16, 3))
def test_MODEL_predictionAugmentated_3D(self): data_aug = Data_Augmentation() pp = Preprocessor(self.data_io3D, batch_size=1, patch_shape=(8, 8, 8), data_aug=data_aug, analysis="patchwise-crop") nn = Neural_Network(preprocessor=pp, architecture=UNet_standard(depth=2)) for sample in self.sample_list3D: predictions = nn.predict_augmentated(sample) self.assertEqual(len(predictions), 3) for pred in predictions: self.assertEqual(pred.shape, (16, 16, 16, 3))