m53297601
/
Paddle
1
0
Fork 0
Code Issues Pull Requests Releases Wiki Activity

refine deformable roi pooling doc (#19944)

Browse Source 
* refine doc, test=develop, test=document_preview
expand_as_op_1
chengjuntao 6 years ago committed by GitHub
parent b1bb23841e
commit 0099e54924
No known key found for this signature in database
GPG Key ID: 4AEE18F83AFDEB23
2 changed files with 75 additions and 33 deletions
Show Stats Download Patch File Download Diff File

2
paddle/fluid/API.spec
Unescape View File

@ -294,7 +294,7 @@ paddle.fluid.layers.where (ArgSpec(args=['condition'], varargs=None, keywords=No
paddle.fluid.layers.sign (ArgSpec(args=['x'], varargs=None, keywords=None, defaults=None), ('document', 'fa2f457a81714430c5677c2d68744728'))
paddle.fluid.layers.deformable_conv (ArgSpec(args=['input', 'offset', 'mask', 'num_filters', 'filter_size', 'stride', 'padding', 'dilation', 'groups', 'deformable_groups', 'im2col_step', 'param_attr', 'bias_attr', 'modulated', 'name'], varargs=None, keywords=None, defaults=(1, 0, 1, None, None, None, None, None, True, None)), ('document', '335193ac57d41d7199f8d26d30c069b1'))
paddle.fluid.layers.unfold (ArgSpec(args=['x', 'kernel_sizes', 'strides', 'paddings', 'dilations', 'name'], varargs=None, keywords=None, defaults=(1, 0, 1, None)), ('document', '3f884662ad443d9ecc2b3734b4f61ad6'))
paddle.fluid.layers.deformable_roi_pooling (ArgSpec(args=['input', 'rois', 'trans', 'no_trans', 'spatial_scale', 'group_size', 'pooled_height', 'pooled_width', 'part_size', 'sample_per_part', 'trans_std', 'position_sensitive', 'name'], varargs=None, keywords=None, defaults=(False, 1.0, [1, 1], 1, 1, None, 1, 0.1, False, None)), ('document', '99c03e3f249e36854f87dedaa17c8f35'))
paddle.fluid.layers.deformable_roi_pooling (ArgSpec(args=['input', 'rois', 'trans', 'no_trans', 'spatial_scale', 'group_size', 'pooled_height', 'pooled_width', 'part_size', 'sample_per_part', 'trans_std', 'position_sensitive', 'name'], varargs=None, keywords=None, defaults=(False, 1.0, [1, 1], 1, 1, None, 1, 0.1, False, None)), ('document', '47c5d1c890b36fa00ff3285c9398f613'))
paddle.fluid.layers.filter_by_instag (ArgSpec(args=['ins', 'ins_tag', 'filter_tag', 'is_lod'], varargs=None, keywords=None, defaults=None), ('document', '7703a2088af8de4128b143ff1164ca4a'))
paddle.fluid.layers.shard_index (ArgSpec(args=['input', 'index_num', 'nshards', 'shard_id', 'ignore_value'], varargs=None, keywords=None, defaults=(-1,)), ('document', 'c4969dd6bf164f9e6a90414ea4f4e5ad'))
paddle.fluid.layers.hard_swish (ArgSpec(args=['x', 'threshold', 'scale', 'offset', 'name'], varargs=None, keywords=None, defaults=(6.0, 6.0, 3.0, None)), ('document', '6a5152a7015c62cb8278fc24cb456459'))

106
python/paddle/fluid/layers/nn.py
Unescape View File

@ -14197,43 +14197,85 @@ def deformable_roi_pooling(input,
position_sensitive=False,
name=None):
"""
Deformable PSROI Pooling Layer
Deformable ROI Pooling Layer
Performs deformable region-of-interest pooling on inputs. As described
in `Deformable Convolutional Networks <https://arxiv.org/abs/1703.06211>`_, it will get offset for each bin after
roi pooling so that pooling at correct region. Batch_size will change to the number of region bounding boxes after deformable_roi_pooling.
The operation has three steps:
Args:
input (Variable):The input of Deformable PSROIPooling.The shape of input tensor is
[N,C,H,W]. Where N is batch size,C is number of input channels,H
is height of the feature, and W is the width of the feature.
rois (Variable): ROIs (Regions of Interest) to pool over.It should be
a 2-D LoDTensor of shape (num_rois, 4), the lod level
is 1. Given as [[x1, y1, x2, y2], ...], (x1, y1) is
the top left coordinates, and (x2, y2) is the bottom
right coordinates.
trans (Variable): Offset of features on ROIs while pooling.The format is NCHW, where
N is number of ROIs, C is number of channels, which indicate the offset distance
in the x and y directions, H is pooled height, and W is pooled width.
no_trans (bool): Whether to add offset to get new value or not while roi pooling, which
value is True or False. Default: False.
spatial_scale (float): Ratio of input feature map height (or width) to raw image height (or width).
Equals the reciprocal of total stride in convolutional layers, Default: 1.0.
group_size (list|tuple): The number of groups which input channels are divided.(eg.number of input channels
is k1*k2*(C+1), which k1 and k2 are group width and height and C+1 is number of output
chanels. eg.(4, 6), which 4 is height of group and 6 is width of group. Default: [1, 1].
pooled_height (integer): The pooled output height. Default: 1.
pooled_width (integer): The pooled output width. Default: 1.
part_size (list|tuple): The height and width of offset, eg.(4, 6), which height is 4 and width is 6, Default:
if None, default value is [pooled_height, pooled_width].
sample_per_part (integer): The number of samples in each bin. Default: 1.
trans_std (float): Coefficient of offset. Default: 0.1.
position_sensitive (bool): Whether to choose deformable psroi pooling mode or not. Default: False.
name (str): Name of layer. Default: None.
Returns:
Variable: The tensor variable storing the deformable psroi pooling \
result.
1. Dividing each region proposal into equal-sized sections with the pooled_width and pooled_height.
2. Add offset to pixel in ROI to get new location and the new value which are computed directly through
bilinear interpolation with four nearest pixel.
3. Sample several points in each bin to get average values as output.
Args:
input (Variable):The input of deformable roi pooling and it is tensor which value type is float32. The shape of input is
[N, C, H, W]. Where N is batch size, C is number of input channels,
H is height of the feature, and W is the width of the feature.
rois (Variable): ROIs (Regions of Interest) with type float32 to pool over. It should be
a 2-D LoDTensor of shape (num_rois, 4), and the lod level
is 1. Given as [[x1, y1, x2, y2], ...], (x1, y1) is
the top left coordinates, and (x2, y2) is the bottom
right coordinates, which value type is float32.
trans (Variable): Offset of features on ROIs while pooling which value type is float32. The format is [N, C, H, W], where
N is number of ROIs, C is number of channels, which indicate the offset distance
in the x and y directions, H is pooled height, and W is pooled width.
no_trans (bool): Whether to add offset to get new value or not while roi pooling, which value with type bool is True or False.
If value is True, no offset will be added in operation. Default: False.
spatial_scale (float): Ratio of input feature map height (or width) to raw image height (or width), which value type is float32.
Equals the reciprocal of total stride in convolutional layers, Default: 1.0.
group_size (list|tuple): The number of groups which input channels are divided and the input is list or tuple, which value type is int32. (eg.number of input channels
is k1 * k2 * (C + 1), which k1 and k2 are group width and height and C+1 is number of output
chanels.) eg.(4, 6), which 4 is height of group and 6 is width of group. Default: [1, 1].
pooled_height (int): The pooled output height which value type is int32. Default: 1.
pooled_width (int): The pooled output width which value type is int32. Default: 1.
part_size (list|tuple): The height and width of offset which values in list or tuple is int32, eg.(4, 6), which height is 4 and width is 6, and values always equal to pooled_height \
and pooled_width. Default: if None, default value is [pooled_height, pooled_width].
sample_per_part (int): The number of samples in each bin which value type is int32. If value is bigger, it will consume more performance. Default: 1.
trans_std (float): Coefficient of offset which value type is float32. It controls weight of offset. Default: 0.1.
position_sensitive (bool): Whether to choose deformable psroi pooling mode or not, and value type is bool(True or False). If value is False, input dimension equals to output dimension. \
If value is True, input dimension shoule be output dimension * pooled_height * pooled_width. Default: False.
name (str|None): Name of layer. Default: None.
Returns:
Variable: Output of deformable roi pooling is that, if position sensitive is False, input dimension equals to output dimension. If position sensitive is True,\
input dimension should be the result of output dimension divided by pooled height and pooled width.
Examples:
.. code-block:: python
# position_sensitive=True
import paddle.fluid as fluid
input = fluid.layers.data(name="input",
shape=[2, 192, 64, 64],
dtype='float32',
append_batch_size=False)
rois = fluid.layers.data(name="rois",
shape=[4],
dtype='float32',
lod_level=1)
trans = fluid.layers.data(name="trans",
shape=[2, 384, 64, 64],
dtype='float32',
append_batch_size=False)
x = fluid.layers.nn.deformable_roi_pooling(input=input,
rois=rois,
trans=trans,
no_trans=False,
spatial_scale=1.0,
group_size=(1, 1),
pooled_height=8,
pooled_width=8,
part_size=(8, 8),
sample_per_part=4,
trans_std=0.1,
position_sensitive=True)
# position_sensitive=False
import paddle.fluid as fluid
input = fluid.layers.data(name="input",
shape=[2, 192, 64, 64],

Write Preview
Loading…
Cancel
Save