Spaces:
Runtime error
Runtime error
File size: 5,571 Bytes
c5b5437 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 |
# copyright (c) 2019 PaddlePaddle Authors. All Rights Reserve.
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
# http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.
from __future__ import absolute_import
from __future__ import division
from __future__ import print_function
import math
import paddle
from paddle import nn
import paddle.nn.functional as F
from paddle import ParamAttr
from ppocr.modeling.backbones.det_mobilenet_v3 import ConvBNLayer
def get_bias_attr(k):
stdv = 1.0 / math.sqrt(k * 1.0)
initializer = paddle.nn.initializer.Uniform(-stdv, stdv)
bias_attr = ParamAttr(initializer=initializer)
return bias_attr
class Head(nn.Layer):
def __init__(self, in_channels, kernel_list=[3, 2, 2], **kwargs):
super(Head, self).__init__()
self.conv1 = nn.Conv2D(
in_channels=in_channels,
out_channels=in_channels // 4,
kernel_size=kernel_list[0],
padding=int(kernel_list[0] // 2),
weight_attr=ParamAttr(),
bias_attr=False)
self.conv_bn1 = nn.BatchNorm(
num_channels=in_channels // 4,
param_attr=ParamAttr(
initializer=paddle.nn.initializer.Constant(value=1.0)),
bias_attr=ParamAttr(
initializer=paddle.nn.initializer.Constant(value=1e-4)),
act='relu')
self.conv2 = nn.Conv2DTranspose(
in_channels=in_channels // 4,
out_channels=in_channels // 4,
kernel_size=kernel_list[1],
stride=2,
weight_attr=ParamAttr(
initializer=paddle.nn.initializer.KaimingUniform()),
bias_attr=get_bias_attr(in_channels // 4))
self.conv_bn2 = nn.BatchNorm(
num_channels=in_channels // 4,
param_attr=ParamAttr(
initializer=paddle.nn.initializer.Constant(value=1.0)),
bias_attr=ParamAttr(
initializer=paddle.nn.initializer.Constant(value=1e-4)),
act="relu")
self.conv3 = nn.Conv2DTranspose(
in_channels=in_channels // 4,
out_channels=1,
kernel_size=kernel_list[2],
stride=2,
weight_attr=ParamAttr(
initializer=paddle.nn.initializer.KaimingUniform()),
bias_attr=get_bias_attr(in_channels // 4), )
def forward(self, x, return_f=False):
x = self.conv1(x)
x = self.conv_bn1(x)
x = self.conv2(x)
x = self.conv_bn2(x)
if return_f is True:
f = x
x = self.conv3(x)
x = F.sigmoid(x)
if return_f is True:
return x, f
return x
class DBHead(nn.Layer):
"""
Differentiable Binarization (DB) for text detection:
see https://arxiv.org/abs/1911.08947
args:
params(dict): super parameters for build DB network
"""
def __init__(self, in_channels, k=50, **kwargs):
super(DBHead, self).__init__()
self.k = k
self.binarize = Head(in_channels, **kwargs)
self.thresh = Head(in_channels, **kwargs)
def step_function(self, x, y):
return paddle.reciprocal(1 + paddle.exp(-self.k * (x - y)))
def forward(self, x, targets=None):
shrink_maps = self.binarize(x)
if not self.training:
return {'maps': shrink_maps}
threshold_maps = self.thresh(x)
binary_maps = self.step_function(shrink_maps, threshold_maps)
y = paddle.concat([shrink_maps, threshold_maps, binary_maps], axis=1)
return {'maps': y}
class LocalModule(nn.Layer):
def __init__(self, in_c, mid_c, use_distance=True):
super(self.__class__, self).__init__()
self.last_3 = ConvBNLayer(in_c + 1, mid_c, 3, 1, 1, act='relu')
self.last_1 = nn.Conv2D(mid_c, 1, 1, 1, 0)
def forward(self, x, init_map, distance_map):
outf = paddle.concat([init_map, x], axis=1)
# last Conv
out = self.last_1(self.last_3(outf))
return out
class PFHeadLocal(DBHead):
def __init__(self, in_channels, k=50, mode='small', **kwargs):
super(PFHeadLocal, self).__init__(in_channels, k, **kwargs)
self.mode = mode
self.up_conv = nn.Upsample(scale_factor=2, mode="nearest", align_mode=1)
if self.mode == 'large':
self.cbn_layer = LocalModule(in_channels // 4, in_channels // 4)
elif self.mode == 'small':
self.cbn_layer = LocalModule(in_channels // 4, in_channels // 8)
def forward(self, x, targets=None):
shrink_maps, f = self.binarize(x, return_f=True)
base_maps = shrink_maps
cbn_maps = self.cbn_layer(self.up_conv(f), shrink_maps, None)
cbn_maps = F.sigmoid(cbn_maps)
if not self.training:
return {'maps': 0.5 * (base_maps + cbn_maps), 'cbn_maps': cbn_maps}
threshold_maps = self.thresh(x)
binary_maps = self.step_function(shrink_maps, threshold_maps)
y = paddle.concat([cbn_maps, threshold_maps, binary_maps], axis=1)
return {'maps': y, 'distance_maps': cbn_maps, 'cbn_maps': binary_maps}
|