summaryrefslogtreecommitdiff
path: root/src/text_recognizer/datasets/transforms.py
diff options
context:
space:
mode:
Diffstat (limited to 'src/text_recognizer/datasets/transforms.py')
-rw-r--r--src/text_recognizer/datasets/transforms.py18
1 files changed, 17 insertions, 1 deletions
diff --git a/src/text_recognizer/datasets/transforms.py b/src/text_recognizer/datasets/transforms.py
index 1ec23dc..016ec80 100644
--- a/src/text_recognizer/datasets/transforms.py
+++ b/src/text_recognizer/datasets/transforms.py
@@ -4,7 +4,7 @@ from PIL import Image
import torch
from torch import Tensor
import torch.nn.functional as F
-from torchvision.transforms import Compose, RandomAffine, ToTensor
+from torchvision.transforms import Compose, RandomAffine, RandomHorizontalFlip, ToTensor
from text_recognizer.datasets.util import EmnistMapper
@@ -77,3 +77,19 @@ class ApplyContrast:
"""Apply mask binary mask to input tensor."""
mask = x > np.random.RandomState().uniform(low=self.low, high=self.high)
return x * mask
+
+
+class Unsqueeze:
+ """Add a dimension to the tensor."""
+
+ def __call__(self, x: Tensor) -> Tensor:
+ """Adds dim."""
+ return x.unsqueeze(0)
+
+
+class Squeeze:
+ """Removes the first dimension of a tensor."""
+
+ def __call__(self, x: Tensor) -> Tensor:
+ """Removes first dim."""
+ return x.squeeze(0)