Add mssim score calculation, corrupt test data before testing, clip noise to avoid invalid values

models/base_dataset.py

@@ -110,3 +110,7 @@ class BaseDataset(Dataset):
         # Save a batch of tensors to a sample file for comparison (no implementation for base dataset)
         pass
 
+    def calculate_score(self, originals, reconstruction, device):
+        # Calculate the score given an uncorrupted and a corrupted batch. (no implementation for base dataset)
+        pass
+

models/base_encoder.py

@@ -6,10 +6,12 @@ import torch
 
 from typing import Optional
 
+from pytorch_msssim import ssim
 from torch.nn.modules.loss import _Loss
 from torchvision.utils import save_image
 
 from config import TRAIN_TEMP_DATA_BASE_PATH, TEST_TEMP_DATA_BASE_PATH, MODEL_STORAGE_BASE_PATH
+from models.base_corruption import BaseCorruption
 from models.base_dataset import BaseDataset
 
 
@@ -175,31 +177,46 @@ class BaseEncoder(torch.nn.Module):
         return losses
 
 
-    def test_encoder(self, dataset: BaseDataset, batch_size: int = 128, num_workers: int = 4):
+    def test_encoder(self, dataset: BaseDataset, corruption: BaseCorruption, batch_size: int = 128, num_workers: int = 4):
         self.log.debug("Getting testing dataset DataLoader.")
         test_loader = dataset.get_test_loader(batch_size=batch_size, num_workers=num_workers)
 
         self.log.debug(f"Start testing...")
+        avg_scores = []
         i = 0
         for batch in test_loader:
             dataset.save_batch_to_sample(
                 batch=batch,
                 filename=os.path.join(TEST_TEMP_DATA_BASE_PATH,
-                                      f'{self.name}_{dataset.name}_test_input_{i}')
+                                      f'{self.name}_{dataset.name}_test_input_{i}_uncorrupted')
+            )
+            corrupted_batch = torch.tensor([corruption.corrupt_image(i) for i in batch], dtype=torch.float32)
+            dataset.save_batch_to_sample(
+                batch=corrupted_batch,
+                filename=os.path.join(TEST_TEMP_DATA_BASE_PATH,
+                                      f'{self.name}_{dataset.name}_test_input_{i}_corrupted')
             )
 
             # load batch features to the active device
-            batch = batch.to(self.device)
+            corrupted_batch = corrupted_batch.to(self.device)
-            outputs = self.process_outputs_for_testing(self(batch))
+            outputs = self.process_outputs_for_testing(self(corrupted_batch))
             img = outputs.cpu().data
             dataset.save_batch_to_sample(
                 batch=img,
                 filename=os.path.join(TEST_TEMP_DATA_BASE_PATH,
                                       f'{self.name}_{dataset.name}_test_reconstruction_{i}')
             )
+
+            batch_score = dataset.calculate_score(batch, img, self.device)
+            avg_scores.append(batch_score)
+
             i += 1
             break
 
+        avg_score = sum(avg_scores) / len(avg_scores)
+        # self.log.warning(f"Testing results - Average score: {avg_score}")
+        print(f"Testing results - Average score: {avg_score}")
+
     def process_loss(self, train_loss, features, outputs) -> _Loss:
         return train_loss
 

models/cifar10_dataset.py

@@ -3,6 +3,7 @@ import os
 from typing import Optional
 
 import numpy
+from pytorch_msssim import ssim
 from torchvision import transforms
 from torchvision.utils import save_image
 
@@ -70,6 +71,28 @@ class Cifar10Dataset(BaseDataset):
 
         return img
 
+    def get_as_image_array(self, item):
+        # Get image data
+        img = self._data[item]
+
+        img = img.reshape((3, 1024))
+
+        # Run transforms
+        if self.transform is not None:
+            img = self.transform(img)
+
+        # Reshape the 32x32x3 image to a 1x3072 array for the Linear layer
+        img = img.view(-1, 3, 32, 32)
+
+        return img
+
     def save_batch_to_sample(self, batch, filename):
         img = batch.view(batch.size(0), 3, 32, 32)
         save_image(img, f"{filename}.png")
+
+    def calculate_score(self, originals, reconstruction, device):
+        # Calculate SSIM
+        originals = originals.view(originals.size(0), 3, 32, 32).to(device)
+        reconstruction = reconstruction.view(reconstruction.size(0), 3, 32, 32).to(device)
+        batch_average_score = ssim(originals, reconstruction, data_range=1, size_average=True)
+        return batch_average_score

models/gaussian_corruption.py

@@ -1,3 +1,4 @@
+import torch
 from torch import Tensor
 
 from models.base_corruption import BaseCorruption
@@ -6,11 +7,13 @@ import numpy
 
 
 def add_noise(image):
+    if isinstance(image, Tensor):
+        image = image.numpy()
     image = image.astype(numpy.float32)
     mean, variance = 0, 0.1
     sigma = variance ** 0.5
     noise = numpy.random.normal(mean, sigma, image.shape).reshape(image.shape)
-    return image + noise
+    return numpy.clip(image + noise, 0, 1)
 
 
 class GaussianCorruption(BaseCorruption):
@@ -25,7 +28,8 @@ class GaussianCorruption(BaseCorruption):
 
     @classmethod
     def corrupt_dataset(cls, dataset: BaseDataset) -> BaseDataset:
-        data = list(map(add_noise, dataset))
+        data = [cls.corrupt_image(x) for x in dataset]
+        # data = list(map(add_noise, dataset._data))
         train_set = cls.corrupt_dataset(dataset.get_train()) if dataset.has_train() else None
         test_set = cls.corrupt_dataset(dataset.get_test()) if dataset.has_test() else None
         return dataset.__class__.get_new(

models/mnist_dataset.py

@@ -2,6 +2,7 @@ import os
 
 from typing import Optional
 
+from pytorch_msssim import ssim
 from torchvision import transforms
 from torchvision.datasets import MNIST
 from torchvision.utils import save_image
@@ -56,3 +57,10 @@ class MNISTDataset(BaseDataset):
     def save_batch_to_sample(self, batch, filename):
         img = batch.view(batch.size(0), 1, 28, 28)
         save_image(img, f"{filename}.png")
+
+    def calculate_score(self, originals, reconstruction, device):
+        # Calculate SSIM
+        originals = originals.view(originals.size(0), 1, 28, 28).to(device)
+        reconstruction = reconstruction.view(reconstruction.size(0), 1, 28, 28).to(device)
+        batch_average_score = ssim(originals, reconstruction, data_range=1, size_average=True)
+        return batch_average_score

models/test_run.py

@@ -59,7 +59,7 @@ class TestRun:
 
         # Test encoder
         self.log.info("Testing auto-encoder...")
-        self.encoder.test_encoder(self.dataset, num_workers=multiprocessing.cpu_count() - 1)
+        self.encoder.test_encoder(self.dataset, corruption=self.corruption, num_workers=multiprocessing.cpu_count() - 1)
 
         self.log.info("Done!")
 

requirements.txt

@@ -3,3 +3,4 @@ torchvision==0.8.2
 torchaudio===0.7.2
 tabulate
 matplotlib
+pytorch-msssim