Merge branch 'master' into prepare-model-param

chainercv/chainer_experimental/datasets/sliceable/getter_dataset.py

@@ -68,17 +68,12 @@ def add_getter(self, keys, getter):
         """Register a getter function
 
         Args:
-            keys (int or string or tuple of strings): The number or name(s) of
-                data that the getter function returns.
+            keys (string or tuple of strings): The name(s) of data
+                that the getter function returns.
             getter (callable): A getter function that takes an index and
                 returns data of the corresponding example.
         """
         self._getters.append(getter)
-        if isinstance(keys, int):
-            if keys == 1:
-                keys = None
-            else:
-                keys = (None,) * keys
         if _is_iterable(keys):
             for key_index, key in enumerate(keys):
                 self._keys.append((key, len(self._getters) - 1, key_index))

chainercv/chainer_experimental/datasets/sliceable/transform_dataset.py

@@ -22,8 +22,8 @@ class TransformDataset(GetterDataset):
     Args:
         dataset: The underlying dataset.
             This dataset should have :meth:`__len__` and :meth:`__getitem__`.
-        keys (int or string or tuple of strings): The number or name(s) of
-            data that the transform function returns.
+        keys (string or tuple of strings): The name(s) of data
+            that the transform function returns.
             If this parameter is omitted, :meth:`__init__` fetches a sample
             from the underlying dataset to determine the number of data.
         transform (callable): A function that is called to transform values
@@ -38,12 +38,7 @@ def __init__(self, dataset, keys, transform=None):
         self._dataset = dataset
         self._transform = transform
 
-        if isinstance(keys, int):
-            if keys == 1:
-                keys = None
-            else:
-                keys = (None,) * keys
-        elif keys is None:
+        if keys is None:
             sample = self._get(0)
             if isinstance(sample, tuple):
                 keys = (None,) * len(sample)

chainercv/visualizations/vis_bbox.py

@@ -37,19 +37,11 @@ def vis_bbox(img, bbox, label=None, score=None, label_names=None,
         >>> plt.show()
 
     Args:
-        img (~numpy.ndarray): An array of shape :math:`(3, height, width)`.
-            This is in RGB format and the range of its value is
-            :math:`[0, 255]`. If this is :obj:`None`, no image is displayed.
-        bbox (~numpy.ndarray): An array of shape :math:`(R, 4)`, where
-            :math:`R` is the number of bounding boxes in the image.
-            Each element is organized
-            by :math:`(y_{min}, x_{min}, y_{max}, x_{max})` in the second axis.
-        label (~numpy.ndarray): An integer array of shape :math:`(R,)`.
-            The values correspond to id for label names stored in
-            :obj:`label_names`. This is optional.
-        score (~numpy.ndarray): A float array of shape :math:`(R,)`.
-             Each value indicates how confident the prediction is.
-             This is optional.
+        img (~numpy.ndarray): See the table below. If this is :obj:`None`,
+            no image is displayed.
+        bbox (~numpy.ndarray): See the table below.
+        label (~numpy.ndarray): See the table below. This is optional.
+        score (~numpy.ndarray): See the table below. This is optional.
         label_names (iterable of strings): Name of labels ordered according
             to label ids. If this is :obj:`None`, labels will be skipped.
         instance_colors (iterable of tuples): List of colors.
@@ -66,6 +58,17 @@ def vis_bbox(img, bbox, label=None, score=None, label_names=None,
         ax (matplotlib.axes.Axis): The visualization is displayed on this
             axis. If this is :obj:`None` (default), a new axis is created.
 
+    .. csv-table::
+        :header: name, shape, dtype, format
+
+        :obj:`img`, ":math:`(3, H, W)`", :obj:`float32`, \
+        "RGB, :math:`[0, 255]`"
+        :obj:`bbox`, ":math:`(R, 4)`", :obj:`float32`, \
+        ":math:`(y_{min}, x_{min}, y_{max}, x_{max})`"
+        :obj:`label`, ":math:`(R,)`", :obj:`int32`, \
+        ":math:`[0, \#fg\_class - 1]`"
+        :obj:`score`, ":math:`(R,)`", :obj:`float32`, --
+
     Returns:
         ~matploblib.axes.Axes:
         Returns the Axes object with the plot for further tweaking.

chainercv/visualizations/vis_image.py

@@ -5,12 +5,17 @@ def vis_image(img, ax=None):
     """Visualize a color image.
 
     Args:
-        img (~numpy.ndarray): An array of shape :math:`(3, height, width)`.
-            This is in RGB format and the range of its value is
-            :math:`[0, 255]`. If this is :obj:`None`, no image is displayed.
+        img (~numpy.ndarray): See the table below.
+            If this is :obj:`None`, no image is displayed.
         ax (matplotlib.axes.Axis): The visualization is displayed on this
             axis. If this is :obj:`None` (default), a new axis is created.
 
+    .. csv-table::
+        :header: name, shape, dtype, format
+
+        :obj:`img`, ":math:`(3, H, W)`", :obj:`float32`, \
+        "RGB, :math:`[0, 255]`"
+
     Returns:
         ~matploblib.axes.Axes:
         Returns the Axes object with the plot for further tweaking.

chainercv/visualizations/vis_instance_segmentation.py

@@ -51,19 +51,11 @@ def vis_instance_segmentation(
         >>> plt.show()
 
     Args:
-        img (~numpy.ndarray): An array of shape :math:`(3, H, W)`.
-            This is in RGB format and the range of its value is
-            :math:`[0, 255]`. If this is :obj:`None`, no image is displayed.
-        mask (~numpy.ndarray): A bool array of shape
-            :math`(R, H, W)`.
-            If there is an object, the value of the pixel is :obj:`True`,
-            and otherwise, it is :obj:`False`.
-        label (~numpy.ndarray): An integer array of shape :math:`(R, )`.
-            The values correspond to id for label names stored in
-            :obj:`label_names`.
-        score (~numpy.ndarray): A float array of shape :math:`(R,)`.
-             Each value indicates how confident the prediction is.
-             This is optional.
+        img (~numpy.ndarray): See the table below. If this is :obj:`None`,
+            no image is displayed.
+        mask (~numpy.ndarray): See the table below.
+        label (~numpy.ndarray): See the table below. This is optional.
+        score (~numpy.ndarray): See the table below. This is optional.
         label_names (iterable of strings): Name of labels ordered according
             to label ids.
         instance_colors (iterable of tuple): List of colors.
@@ -82,6 +74,16 @@ def vis_instance_segmentation(
         ax (matplotlib.axes.Axis): The visualization is displayed on this
             axis. If this is :obj:`None` (default), a new axis is created.
 
+    .. csv-table::
+        :header: name, shape, dtype, format
+
+        :obj:`img`, ":math:`(3, H, W)`", :obj:`float32`, \
+        "RGB, :math:`[0, 255]`"
+        :obj:`mask`, ":math:`(R, H, W)`", :obj:`bool`, --
+        :obj:`label`, ":math:`(R,)`", :obj:`int32`, \
+        ":math:`[0, \#fg\_class - 1]`"
+        :obj:`score`, ":math:`(R,)`", :obj:`float32`, --
+
     Returns:
         matploblib.axes.Axes: Returns :obj:`ax`.
         :obj:`ax` is an :class:`matploblib.axes.Axes` with the plot.

chainercv/visualizations/vis_semantic_segmentation.py

@@ -32,13 +32,9 @@ def vis_semantic_segmentation(
         >>> plt.show()
 
     Args:
-        img (~numpy.ndarray): An array of shape :math:`(3, height, width)`.
-            This is in RGB format and the range of its value is
-            :math:`[0, 255]`. If this is :obj:`None`, no image is displayed.
-        label (~numpy.ndarray): An integer array of shape
-            :math:`(height, width)`.
-            The values correspond to id for label names stored in
-            :obj:`label_names`.
+        img (~numpy.ndarray): See the table below. If this is :obj:`None`,
+            no image is displayed.
+        label (~numpy.ndarray): See the table below.
         label_names (iterable of strings): Name of labels ordered according
             to label ids.
         label_colors: (iterable of tuple): An iterable of colors for regular
@@ -63,6 +59,14 @@ def vis_semantic_segmentation(
         ax (matplotlib.axes.Axis): The visualization is displayed on this
             axis. If this is :obj:`None` (default), a new axis is created.
 
+    .. csv-table::
+        :header: name, shape, dtype, format
+
+        :obj:`img`, ":math:`(3, H, W)`", :obj:`float32`, \
+        "RGB, :math:`[0, 255]`"
+        :obj:`label`, ":math:`(H, W)`", :obj:`int32`, \
+        ":math:`[-1, \#class - 1]`"
+
     Returns:
         matploblib.axes.Axes and list of matplotlib.patches.Patch:
         Returns :obj:`ax` and :obj:`legend_handles`.

examples/deeplab/README.md

@@ -16,9 +16,9 @@ So we evaluated the pretrained graph using `eval_semantic_segmentation` in Chain
 Although we could generate a graph the biggest image can be input, it resulted 40.13% in official evaluation code.
 
 ## Demo
-This demo downloads Cityscapes pretrained model automatically if a pretrained model path is not given.
+This demo downloads a pretrained model automatically if a pretrained model path is not given.
 ```
-$ python demo.py [--gpu <gpu>] [--pretrained-model <model_path>] [--input-size <size>] <image>.jpg
+$ python demo.py [--dataset cityscapes|ade20k|voc] [--gpu <gpu>] [--pretrained-model <model_path>] [--min-input-size <size>] <image>.jpg
 ```
 
 

examples/deeplab/demo.py

@@ -1,5 +1,5 @@
 import argparse
-
+import copy
 import matplotlib.pyplot as plt
 
 import chainer
@@ -19,7 +19,8 @@
 def main():
     parser = argparse.ArgumentParser()
     parser.add_argument('--gpu', type=int, default=-1)
-    parser.add_argument('--pretrained-model', default=None)
+    parser.add_argument('--pretrained-model')
+    parser.add_argument('--min-input-size', type=int, default=None)
     parser.add_argument(
         '--dataset', choices=('cityscapes', 'ade20k', 'voc'),
         default='cityscapes')
@@ -42,9 +43,10 @@ def main():
         label_names = voc_semantic_segmentation_label_names
         colors = voc_semantic_segmentation_label_colors
 
+    params = copy.deepcopy(DeepLabV3plusXception65.preset_params[args.dataset])
+    params['min_input_size'] = args.min_input_size
     model = DeepLabV3plusXception65(
-        pretrained_model=args.pretrained_model,
-        **DeepLabV3plusXception65.preset_params[args.dataset])
+        pretrained_model=args.pretrained_model, **params)
 
     if args.gpu >= 0:
         chainer.cuda.get_device_from_id(args.gpu).use()
@@ -59,7 +61,7 @@ def main():
     vis_image(img, ax=ax1)
     ax2 = fig.add_subplot(1, 2, 2)
     # Do not overlay the label image on the color image
-    vis_semantic_segmentation(
+    vis_semantic_segmentation(>>>>>>> master
         None, label, label_names, colors, ax=ax2)
     plt.show()
 

examples/faster_rcnn/demo.py

@@ -14,7 +14,7 @@ def main():
     parser.add_argument('--gpu', type=int, default=-1)
     parser.add_argument('--pretrained-model')
     parser.add_argument(
-        '--dataset', choices=('voc'), default='voc')
+        '--dataset', choices=('voc',), default='voc')
     parser.add_argument('image')
     args = parser.parse_args()
 

examples/fcis/README.md

@@ -20,7 +20,7 @@
 Segment objects in an given image. This demo downloads SBD pretrained model automatically if a pretrained model path is not given.
 
 ```bash
-python demo.py [--gpu <gpu>] [--pretrained-model <model_path>] [--dataset <sbd, coco>] <image.jpg>
+python demo.py [--dataset sbd|coco] [--gpu <gpu>] [--pretrained-model <model_path>] <image.jpg>
 ```
 
 ## Evaluation

examples/fpn/demo.py

@@ -24,7 +24,7 @@ def main():
     parser.add_argument('--gpu', type=int, default=-1)
     parser.add_argument('--pretrained-model')
     parser.add_argument(
-        '--dataset', choices=('coco'), default='coco')
+        '--dataset', choices=('coco',), default='coco')
     parser.add_argument('image')
     args = parser.parse_args()
 

examples/pspnet/README.md

@@ -3,7 +3,7 @@
 ## Demo
 This demo downloads a pretrained model automatically if a pretrained model path is not given.
 ```
-$ python demo.py [--gpu <gpu>] [--pretrained-model <model_path>] [--input-size <size>] <image>.jpg
+$ python demo.py [--dataset cityscapes|ade20k] [--gpu <gpu>] [--pretrained-model <model_path>] [--input-size <size>] <image>.jpg
 ```
 
 ## Weight Covnersion

examples/pspnet/demo.py

@@ -1,8 +1,9 @@
 import argparse
-
+import copy
 import matplotlib.pyplot as plt
 
 import chainer
+
 from chainercv.datasets import ade20k_semantic_segmentation_label_colors
 from chainercv.datasets import ade20k_semantic_segmentation_label_names
 from chainercv.datasets import cityscapes_semantic_segmentation_label_colors
@@ -17,6 +18,7 @@ def main():
     parser = argparse.ArgumentParser()
     parser.add_argument('--gpu', '-g', type=int, default=-1)
     parser.add_argument('--pretrained-model')
+    parser.add_argument('--input-size', type=int, default=713)
     parser.add_argument(
         '--dataset', choices=('cityscapes', 'ade20k'), default='cityscapes')
     parser.add_argument('image')
@@ -33,9 +35,11 @@ def main():
         label_names = ade20k_semantic_segmentation_label_names
         colors = ade20k_semantic_segmentation_label_colors
 
+    input_size = (args.input_size, args.input_size)
+    params = copy.deepcopy(PSPNetResNet101.preset_params[args.dataset])
+    params['input_size'] = input_size
     model = PSPNetResNet101(
-        pretrained_model=args.pretrained_model,
-        **PSPNetResNet101.preset_params[args.dataset])
+        pretrained_model=args.pretrained_model, **params)
 
     if args.gpu >= 0:
         chainer.cuda.get_device_from_id(args.gpu).use()

examples/segnet/demo.py

@@ -17,7 +17,7 @@ def main():
     parser = argparse.ArgumentParser()
     parser.add_argument('--gpu', type=int, default=-1)
     parser.add_argument('--pretrained-model')
-    parser.add_argument('--dataset', choices=('camvid'), default='camvid')
+    parser.add_argument('--dataset', choices=('camvid',), default='camvid')
     parser.add_argument('image')
     args = parser.parse_args()
 

examples/ssd/demo.py

@@ -17,7 +17,7 @@ def main():
     parser.add_argument('--gpu', type=int, default=-1)
     parser.add_argument('--pretrained-model')
     parser.add_argument(
-        '--dataset', choices=('voc'), default='voc')
+        '--dataset', choices=('voc',), default='voc')
     parser.add_argument('image')
     args = parser.parse_args()
 

examples/yolo/demo.py

@@ -19,21 +19,21 @@ def main():
     parser.add_argument('--gpu', type=int, default=-1)
     parser.add_argument('--pretrained-model')
     parser.add_argument(
-        '--dataset', choices=('voc'), default='voc')
+        '--dataset', choices=('voc',), default='voc')
     parser.add_argument('image')
     args = parser.parse_args()
 
     if args.model == 'yolo_v2':
         cls = YOLOv2
     elif args.model == 'yolo_v2_tiny':
-        model = YOLOv2Tiny
+        cls = YOLOv2Tiny
     elif args.model == 'yolo_v3':
-        model = YOLOv3
+        cls = YOLOv3
 
     if args.dataset == 'voc':
         if args.pretrained_model is None:
             args.pretrained_model = 'voc0712'
-        label_names = voc_bbox_label_names
+        label_names = voc_bbox_label_name
 
     model = cls(pretrained_model=args.pretrained_model,
                 **cls.preset_params[args.dataset])

tests/chainer_experimental_tests/datasets_tests/sliceable_tests/test_getter_dataset.py

@@ -11,7 +11,7 @@ def __init__(self, iterable=tuple):
 
         self.add_getter('item0', self.get_item0)
         self.add_getter(iterable(('item1', 'item2')), self.get_item1_item2)
-        self.add_getter(1, self.get_item3)
+        self.add_getter(('item3',), self.get_item3)
 
         self.count = 0
 
@@ -28,7 +28,7 @@ def get_item1_item2(self, i):
 
     def get_item3(self, i):
         self.count += 1
-        return 'item3({:d})'.format(i)
+        return ('item3({:d})'.format(i),)
 
 
 @testing.parameterize(
@@ -43,7 +43,7 @@ def setUp(self):
 
     def test_keys(self):
         self.assertEqual(
-            self.dataset.keys, ('item0', 'item1', 'item2', None))
+            self.dataset.keys, ('item0', 'item1', 'item2', 'item3'))
 
     def test_get_example_by_keys(self):
         example = self.dataset.get_example_by_keys(1, (1, 2, 3))

tests/chainer_experimental_tests/datasets_tests/sliceable_tests/test_transform_dataset.py

@@ -65,20 +65,6 @@ def test_transform(self):
         dataset = TransformDataset(self.dataset, keys, self.func)
         self._check(dataset, self.keys)
 
-    def test_transform_with_n_keys(self):
-        if isinstance(self.keys, tuple):
-            n_keys = len(self.keys)
-            if n_keys == 1:
-                self.skipTest(
-                    'tuple of single element is not supported '
-                    'when the number of keys is specified')
-            expected_keys = (None,) * n_keys
-        else:
-            n_keys = 1
-            expected_keys = None
-        dataset = TransformDataset(self.dataset, n_keys, self.func)
-        self._check(dataset, expected_keys)
-
     def test_transform_compat(self):
         if isinstance(self.keys, tuple):
             expected_keys = (None,) * len(self.keys)