assert_is_point_dataset

chainercv/utils/__init__.py

@@ -15,6 +15,7 @@
 from chainercv.utils.testing import assert_is_image  # NOQA
 from chainercv.utils.testing import assert_is_label_dataset  # NOQA
 from chainercv.utils.testing import assert_is_point  # NOQA
+from chainercv.utils.testing import assert_is_point_dataset  # NOQA
 from chainercv.utils.testing import assert_is_semantic_segmentation_dataset  # NOQA
 from chainercv.utils.testing import assert_is_semantic_segmentation_link  # NOQA
 from chainercv.utils.testing import ConstantStubLink  # NOQA

chainercv/utils/testing/__init__.py

@@ -4,6 +4,7 @@
 from chainercv.utils.testing.assertions import assert_is_image  # NOQA
 from chainercv.utils.testing.assertions import assert_is_label_dataset  # NOQA
 from chainercv.utils.testing.assertions import assert_is_point  # NOQA
+from chainercv.utils.testing.assertions import assert_is_point_dataset  # NOQA
 from chainercv.utils.testing.assertions import assert_is_semantic_segmentation_dataset  # NOQA
 from chainercv.utils.testing.assertions import assert_is_semantic_segmentation_link  # NOQA
 from chainercv.utils.testing.constant_stub_link import ConstantStubLink  # NOQA

chainercv/utils/testing/assertions/__init__.py

@@ -4,5 +4,6 @@
 from chainercv.utils.testing.assertions.assert_is_image import assert_is_image  # NOQA
 from chainercv.utils.testing.assertions.assert_is_label_dataset import assert_is_label_dataset  # NOQA
 from chainercv.utils.testing.assertions.assert_is_point import assert_is_point  # NOQA
+from chainercv.utils.testing.assertions.assert_is_point_dataset import assert_is_point_dataset  # NOQA
 from chainercv.utils.testing.assertions.assert_is_semantic_segmentation_dataset import assert_is_semantic_segmentation_dataset  # NOQA
 from chainercv.utils.testing.assertions.assert_is_semantic_segmentation_link import assert_is_semantic_segmentation_link  # NOQA

chainercv/utils/testing/assertions/assert_is_point_dataset.py

@@ -0,0 +1,59 @@
+import numpy as np
+import six
+
+from chainercv.utils.testing.assertions.assert_is_image import assert_is_image
+from chainercv.utils.testing.assertions.assert_is_point import assert_is_point
+
+
+def assert_is_point_dataset(dataset, n_point=None, n_example=None,
+                            no_mask=False):
+    """Checks if a dataset satisfies the point dataset API.
+
+    This function checks if a given dataset satisfies the point dataset
+    API or not.
+    If the dataset does not satifiy the API, this function raises an
+    :class:`AssertionError`.
+
+    Args:
+        dataset: A dataset to be checked.
+        n_point (int): The number of expected points per image.
+            If thsi is :obj:`None`, the number of points per image can be
+            arbitrary.
+        n_example (int): The number of examples to be checked.
+            If this argument is specified, this function picks
+            examples ramdomly and checks them. Otherwise,
+            this function checks all examples.
+        no_mask (bool): If :obj:`True`, we assume that
+            point mask is always not contained.
+            If :obj:`False`, point mask may or may not be contained.
+
+    """
+
+    assert len(dataset) > 0, 'The length of dataset must be greater than zero.'
+
+    if n_example:
+        for _ in six.moves.range(n_example):
+            i = np.random.randint(0, len(dataset))
+            _check_example(dataset[i], n_point, no_mask)
+    else:
+        for i in six.moves.range(len(dataset)):
+            _check_example(dataset[i], n_point, no_mask)
+
+
+def _check_example(example, n_point=None, no_mask=False):
+    assert len(example) >= 2, \
+        'Each example must have at least two elements:' \
+        'img, point (mask is optional).'
+
+    if len(example) == 2 or no_mask:
+        img, point = example[:2]
+        mask = None
+    elif len(example) >= 3:
+        img, point, mask = example[:3]
+
+    assert_is_image(img, color=True)
+    assert_is_point(point, mask, img.shape[1:])
+
+    if n_point is not None:
+        assert point.shape[0] == n_point, \
+            'The number of points is different from the expected number.'

docs/source/reference/utils.rst

@@ -91,6 +91,10 @@ assert_is_point
 ~~~~~~~~~~~~~~~
 .. autofunction:: assert_is_point
 
+assert_is_point_dataset
+~~~~~~~~~~~~~~~~~~~~~~~
+.. autofunction:: assert_is_point_dataset
+
 assert_is_semantic_segmentation_dataset
 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
 .. autofunction:: assert_is_semantic_segmentation_dataset

tests/utils_tests/testing_tests/assertions_tests/test_assert_is_point_dataset.py

@@ -0,0 +1,106 @@
+import numpy as np
+import unittest
+
+from chainer.dataset import DatasetMixin
+from chainer import testing
+
+from chainercv.utils import assert_is_point_dataset
+
+
+class PointDataset(DatasetMixin):
+
+    H = 48
+    W = 64
+
+    def __init__(self, n_point_candidates,
+                 return_mask, *options):
+        self.n_point_candidates = n_point_candidates
+        self.return_mask = return_mask
+        self.options = options
+
+    def __len__(self):
+        return 10
+
+    def get_example(self, i):
+        img = np.random.randint(0, 256, size=(3, self.H, self.W))
+        n_point = np.random.choice(self.n_point_candidates)
+        point_y = np.random.uniform(0, self.H, size=(n_point,))
+        point_x = np.random.uniform(0, self.W, size=(n_point,))
+        point = np.stack((point_y, point_x), axis=1).astype(np.float32)
+        if self.return_mask:
+            mask = np.random.randint(0, 2, size=(n_point,)).astype(np.bool)
+            return (img, point, mask) + self.options
+        else:
+            return (img, point) + self.options
+
+
+class InvalidSampleSizeDataset(PointDataset):
+
+    def get_example(self, i):
+        img = super(
+            InvalidSampleSizeDataset, self).get_example(i)[0]
+        return img
+
+
+class InvalidImageDataset(PointDataset):
+
+    def get_example(self, i):
+        img = super(
+            InvalidImageDataset, self).get_example(i)[0]
+        rest = super(
+            InvalidImageDataset, self).get_example(i)[1:]
+        return (img[0],) + rest
+
+
+class InvalidPointDataset(PointDataset):
+
+    def get_example(self, i):
+        img, point = super(InvalidPointDataset, self).get_example(i)[:2]
+        rest = super(InvalidPointDataset, self).get_example(i)[2:]
+        point += 1000
+        return (img, point) + rest
+
+
+@testing.parameterize(
+    # No optional Values
+    {'dataset': PointDataset([10, 15], True), 'valid': True, 'n_point': None},
+    {'dataset': PointDataset([10, 15], False), 'valid': True, 'n_point': None},
+    {'dataset': PointDataset([15], True), 'valid': True, 'n_point': 15},
+    {'dataset': PointDataset([15], False), 'valid': True, 'n_point': 15},
+    # Invalid n_point
+    {'dataset': PointDataset([15], True), 'valid': False, 'n_point': 10},
+    {'dataset': PointDataset([15], False), 'valid': False, 'n_point': 10},
+    # Return optional values
+    {'dataset': PointDataset([10, 15], True, 'option'),
+     'valid': True, 'n_point': None},
+    {'dataset': PointDataset([10, 15], False, 'option'),
+     'valid': True, 'n_point': None, 'no_mask': True},
+    {'dataset': PointDataset([15], True, 'option'),
+     'valid': True, 'n_point': 15},
+    {'dataset': PointDataset([15], False, 'option'),
+     'valid': True, 'n_point': 15, 'no_mask': True},
+    # Invalid datasets
+    {'dataset': InvalidSampleSizeDataset([10], True),
+     'valid': False, 'n_point': None},
+    {'dataset': InvalidImageDataset([10], True),
+     'valid': False, 'n_point': None},
+    {'dataset': InvalidPointDataset([10], True),
+     'valid': False, 'n_point': None},
+)
+class TestAssertIsPointDataset(unittest.TestCase):
+
+    def setUp(self):
+        if not hasattr(self, 'no_mask'):
+            self.no_mask = False
+
+    def test_assert_is_point_dataset(self):
+        if self.valid:
+            assert_is_point_dataset(
+                self.dataset, self.n_point, 20, self.no_mask)
+        else:
+            with self.assertRaises(AssertionError):
+                assert_is_point_dataset(
+                    self.dataset, self.n_point, 20, self.no_mask)
+
+
+testing.run_module(__name__, __file__)