fixed augmentations, uploaded data

config/datasets/anomaly_detection_images.yaml

@@ -7,22 +7,19 @@ The MVTec anomaly detection dataset (MVTec AD)
 https://www.mvtec.com/company/research/datasets/mvtec-ad
 DOI: 10.1007/s11263-020-01400-4
 DOI: 10.1109/CVPR.2019.00982
-"
-https://www:
-  mvtec:
-    com/company/research/datasets/mvtec-ad:
+https://www:mvtec:com/company/research/datasets/mvtec-ad"
 markup_info: 'Train images do not contain anomalies'
 date_time: 20.07.2024
 
 _target_: innofw.core.datamodules.lightning_datamodules.anomaly_detection_images.ImageAnomaliesLightningDataModule
 
 train:
-  source: ./data/MVTEC/train
+  source: https://api.blackhole.ai.innopolis.university/public-datasets/anomaly_detection_mvtec/train.zip
   target: ./data/MVTEC/train
 test:
-  source: ./data/MVTEC/test
+  source: https://api.blackhole.ai.innopolis.university/public-datasets/anomaly_detection_mvtec/test.zip
   target: ./data/MVTEC/test
 
 infer:
-  source: ./data/MVTEC/infer
-  target: ./data/MVTEC/infer
+  source: https://api.blackhole.ai.innopolis.university/public-datasets/anomaly_detection_mvtec/test.zip
+  target: ./data/MVTEC/test

config/experiments/anomaly-detection/KG_210724_ba083ak_anomaly_detection_images.yaml

@@ -14,4 +14,11 @@ task: "anomaly-detection-images"
 random_seed: 0
 epochs: 3
 batch_size: 8
-accelerator: cpu # gpu
+accelerator: gpu
+
+wandb:
+  enable: True
+  project: anomaly_detect_mvtec
+  entity: "k-galliamov"
+  group: none
+  job_type: training

innofw/core/datamodules/lightning_datamodules/anomaly_detection_images.py

@@ -44,7 +44,7 @@ def __init__(
             test,
             infer=None,
             batch_size: int = 2,
-            val_size: float = 0.2,
+            val_size: float = 0.5,
             num_workers: int = 1,
             augmentations=None,
             stage=None,
@@ -58,8 +58,10 @@ def __init__(
         self.val_size = val_size
 
     def setup_train_test_val(self, **kwargs):
-        self.train_dataset = AnomaliesDataset(self.train_source, self.aug, add_labels=False)
-        self.test_dataset = AnomaliesDataset(self.test_source, add_labels=True)
+        self.train_dataset = AnomaliesDataset(self.train_source, self.get_aug(self.aug, 'train'),
+                                              add_labels=False)
+        self.test_dataset = AnomaliesDataset(self.test_source, self.get_aug(self.aug, 'test'),
+                                             add_labels=True)
 
         # divide into train, val, test - val is a part of test since train does not have anomalies
         n = len(self.test_dataset)
@@ -101,7 +103,7 @@ def predict_dataloader(self):
         return test_dataloader
 
     def setup_infer(self):
-        self.predict_dataset = AnomaliesDataset(self.predict_source)
+        self.predict_dataset = AnomaliesDataset(self.predict_source, self.aug)
 
     def save_preds(self, preds, stage: Stages, dst_path: pathlib.Path):
         dst_path = pathlib.Path(dst_path)

innofw/core/datasets/anomalies.py

@@ -26,11 +26,11 @@ class AnomaliesDataset(Dataset):
     """
 
     def __init__(self, data_path, augmentations, add_labels=False):
-        self.images = list(Path(data_path + '/images').iterdir())
+        self.images = list(Path(str(data_path) + '/images').iterdir())
         self.add_labels = add_labels
         self.augmentations = augmentations
         if self.add_labels:
-            self.labels = list(Path(data_path + '/labels').iterdir())
+            self.labels = list(Path(str(data_path) + '/labels').iterdir())
 
     def __len__(self):
         return len(self.images)
@@ -39,11 +39,12 @@ def __getitem__(self, idx):
         imagePath = self.images[idx]
         image = cv2.imread(str(imagePath))
         image = cv2.cvtColor(image, cv2.COLOR_BGR2RGB)
-
+        image = torch.from_numpy(image).float()
+        image = torch.div(image, 255)
         if not self.add_labels:
-            return image
+            return self.augmentations(image) if self.augmentations is not None else image
         mask = cv2.imread(str(self.labels[idx]), 0)
-        image, mask = self.augmentations(image, mask)
-        mask = mask[None, :]
+        if self.augmentations is not None:
+            image, mask = self.augmentations(image=image, mask=mask)
 
         return image, mask

innofw/core/models/torch/lightning_modules/anomaly_detection_images.py

@@ -72,7 +72,7 @@ def training_step(self, x, batch_idx):
         loss = self.loss_fn(x, x_rec)
         metrics = self.compute_metrics('train', x_rec, x)
         self.log_metrics('train', metrics)
-        self.log("loss", loss, on_step=False, on_epoch=True)
+        self.log("train_loss", loss, on_step=False, on_epoch=True)
         return {"loss": loss}
 
     def validation_step(self, batch, batch_idx):
@@ -82,7 +82,7 @@ def validation_step(self, batch, batch_idx):
         mask = self.compute_anomaly_mask(x)
         metrics = self.compute_metrics('val', mask, y)
         self.log_metrics('val', metrics)
-        self.log("loss", loss, on_step=False, on_epoch=True)
+        self.log("val_loss", loss, on_step=False, on_epoch=True)
         return {"loss": loss}
 
     def test_step(self, x, batch_idx):
@@ -91,7 +91,7 @@ def test_step(self, x, batch_idx):
         mask = self.compute_anomaly_mask(x)
         metrics = self.compute_metrics('val', mask, y)
         self.log_metrics('val', metrics)
-        self.log("loss", loss, on_step=False, on_epoch=True)
+        self.log("test_loss", loss, on_step=False, on_epoch=True)
         return {"loss": loss}
 
     def predict_step(self, x, batch_idx, **kwargs):