summaryrefslogtreecommitdiff
path: root/text_recognizer/data/transforms.py
diff options
context:
space:
mode:
authorGustaf Rydholm <gustaf.rydholm@gmail.com>2021-04-03 21:59:07 +0200
committerGustaf Rydholm <gustaf.rydholm@gmail.com>2021-04-03 21:59:07 +0200
commit07f2cc3665a1a60efe8ed8073cad6ac4f344b2c2 (patch)
treed24ae8e3b9b39bfcfb3b850b30cb966eb3b064a7 /text_recognizer/data/transforms.py
parent3196144ec99e803cef218295ddea592748931c57 (diff)
Add IAM paragraphs dataset
Diffstat (limited to 'text_recognizer/data/transforms.py')
-rw-r--r--text_recognizer/data/transforms.py11
1 files changed, 4 insertions, 7 deletions
diff --git a/text_recognizer/data/transforms.py b/text_recognizer/data/transforms.py
index 2291eec..616e236 100644
--- a/text_recognizer/data/transforms.py
+++ b/text_recognizer/data/transforms.py
@@ -9,7 +9,8 @@ from torch import Tensor
from text_recognizer.datasets.iam_preprocessor import Preprocessor
from text_recognizer.data.emnist import emnist_mapping
-
+
+
class ToLower:
"""Converts target to lower case."""
@@ -23,15 +24,11 @@ class ToCharcters:
"""Converts integers to characters."""
def __init__(self) -> None:
- self.mapping, _, _ = emnist_mapping()
+ self.mapping, _, _ = emnist_mapping()
def __call__(self, y: Tensor) -> str:
"""Converts a Tensor to a str."""
- return (
- "".join([self.mapping(int(i)) for i in y])
- .strip("<p>")
- .replace(" ", "▁")
- )
+ return "".join([self.mapping(int(i)) for i in y]).strip("<p>").replace(" ", "▁")
class WordPieces: