apply changes

mindee · fg-mindee · Nov 30, 2021 · Nov 13, 2021 · Nov 15, 2021 · Nov 15, 2021
commit 247ef9215f734f3e6b80588d02c8d70adf90fc08
diff --git a/docs/source/datasets.rst b/docs/source/datasets.rst
@@ -22,7 +22,7 @@ Here are all datasets that are available through docTR:
 .. autoclass:: IIIT5K
 .. autoclass:: SVT
 .. autoclass:: SynthText
-.. autoclass:: ICDAR2003
+.. autoclass:: IC03
 
 
 Data Loading

diff --git a/doctr/datasets/__init__.py b/doctr/datasets/__init__.py
@@ -5,7 +5,7 @@
 from .detection import *
 from .doc_artefacts import *
 from .funsd import *
-from .icdar2003 import *
+from .ic03 import *
 from .iiit5k import *
 from .ocr import *
 from .recognition import *

diff --git a/doctr/datasets/icdar2003.py → doctr/datasets/ic03.py b/doctr/datasets/icdar2003.py → doctr/datasets/ic03.py
@@ -11,16 +11,16 @@
 
 from .datasets import VisionDataset
 
-__all__ = ['ICDAR2003']
+__all__ = ['IC03']
 
 
-class ICDAR2003(VisionDataset):
-    """ICDAR2003 dataset from `"ICDAR 2003 Robust Reading Competitions: Entries, Results and Future Directions"
+class IC03(VisionDataset):
+    """IC03 dataset from `"ICDAR 2003 Robust Reading Competitions: Entries, Results and Future Directions"
     <http://www.iapr-tc11.org/mediawiki/index.php?title=ICDAR_2003_Robust_Reading_Competitions>`_.
 
     Example::
-        >>> from doctr.datasets import ICDAR2003
-        >>> train_set = ICDAR2003(train=True, download=True)
+        >>> from doctr.datasets import IC03
+        >>> train_set = IC03(train=True, download=True)
         >>> img, target = train_set[0]
 
     Args:
@@ -82,7 +82,7 @@ def __init__(
                 ]
 
             # filter images without boxes
-            if _boxes:
+            if len(_boxes) > 0:
                 # Convert them to relative
                 w, h = int(resolution.attrib['x']), int(resolution.attrib['y'])
                 boxes = np.asarray(_boxes, dtype=np_dtype)

diff --git a/doctr/datasets/svt.py b/doctr/datasets/svt.py
@@ -61,13 +61,15 @@ def __init__(
                 raise FileNotFoundError(f"unable to locate {os.path.join(tmp_root, name.text)}")
 
             if rotated_bbox:
+                # x_center, y_center, width, height, 0
                 _boxes = [
                     [float(rect.attrib['x']) + float(rect.attrib['width']) / 2,
                      float(rect.attrib['y']) + float(rect.attrib['height']) / 2,
-                     float(rect.attrib['width']), float(rect.attrib['height'])]
+                     float(rect.attrib['width']), float(rect.attrib['height']), 0.0]
                     for rect in rectangles
                 ]
             else:
+                # x_min, y_min, x_max, y_max
                 _boxes = [
                     [float(rect.attrib['x']), float(rect.attrib['y']),
                      float(rect.attrib['x']) + float(rect.attrib['width']),

diff --git a/tests/pytorch/test_datasets_pt.py b/tests/pytorch/test_datasets_pt.py
@@ -34,8 +34,8 @@ def test_visiondataset():
         ['SVT', False, [512, 512], 249, False],
         ['SynthText', True, [512, 512], 27, True],  # Actual set has 772875 samples
         ['SynthText', False, [512, 512], 3, False],  # Actual set has 85875 samples
-        ['ICDAR2003', True, [512, 512], 246, True],
-        ['ICDAR2003', False, [512, 512], 249, False],
+        ['IC03', True, [512, 512], 246, True],
+        ['IC03', False, [512, 512], 249, False],
     ],
 )
 def test_dataset(dataset_name, train, input_size, size, rotate):

diff --git a/tests/tensorflow/test_datasets_tf.py b/tests/tensorflow/test_datasets_tf.py
@@ -24,8 +24,8 @@
         ['SVT', False, [512, 512], 249, False],
         ['SynthText', True, [512, 512], 27, True],  # Actual set has 772875 samples
         ['SynthText', False, [512, 512], 3, False],  # Actual set has 85875 samples
-        ['ICDAR2003', True, [512, 512], 246, True],
-        ['ICDAR2003', False, [512, 512], 249, False],
+        ['IC03', True, [512, 512], 246, True],
+        ['IC03', False, [512, 512], 249, False],
     ],
 )
 def test_dataset(dataset_name, train, input_size, size, rotate):