Python nn.AffineChannel2d() Examples
The following are 30
code examples of nn.AffineChannel2d().
You can vote up the ones you like or vote down the ones you don't like,
and go to the original project or source file by following the links above each example.
You may also want to check out all available functions/classes of the module
nn
, or try the search function
.
.
Example #1
| Source File: ResNet.py From Large-Scale-VRD.pytorch with MIT License | 5 votes |
|
def basic_bn_shortcut(inplanes, outplanes, stride):
return nn.Sequential(
nn.Conv2d(inplanes,
outplanes,
kernel_size=1,
stride=stride,
bias=False),
mynn.AffineChannel2d(outplanes),
)
Example #2
| Source File: ResNet.py From PANet with MIT License | 5 votes |
|
def _init_modules(self):
assert cfg.RESNETS.FREEZE_AT in [0, 2, 3, 4, 5]
assert cfg.RESNETS.FREEZE_AT <= self.convX
for i in range(1, cfg.RESNETS.FREEZE_AT + 1):
freeze_params(getattr(self, 'res%d' % i))
# Freeze all bn (affine) layers !!!
self.apply(lambda m: freeze_params(m) if isinstance(m, mynn.AffineChannel2d) else None)
Example #3
| Source File: ResNet.py From PANet with MIT License | 5 votes |
|
def basic_bn_shortcut(inplanes, outplanes, stride):
return nn.Sequential(
nn.Conv2d(inplanes,
outplanes,
kernel_size=1,
stride=stride,
bias=False),
mynn.AffineChannel2d(outplanes),
)
Example #4
| Source File: ResNet.py From PANet with MIT License | 5 votes |
|
def basic_bn_stem():
return nn.Sequential(OrderedDict([
('conv1', nn.Conv2d(3, 64, 7, stride=2, padding=3, bias=False)),
('bn1', mynn.AffineChannel2d(64)),
('relu', nn.ReLU(inplace=True)),
# ('maxpool', nn.MaxPool2d(kernel_size=3, stride=2, padding=0, ceil_mode=True))]))
('maxpool', nn.MaxPool2d(kernel_size=3, stride=2, padding=1))]))
Example #5
| Source File: mask_rcnn_heads.py From PANet with MIT License | 5 votes |
|
def __init__(self, dim_in, roi_xform_func, spatial_scale):
super().__init__()
self.dim_in = dim_in
self.roi_xform = roi_xform_func
self.spatial_scale = spatial_scale
self.dim_out = cfg.MRCNN.DIM_REDUCED
self.res5, dim_out = ResNet_roi_conv5_head_for_masks(dim_in)
self.upconv5 = nn.ConvTranspose2d(dim_out, self.dim_out, 2, 2, 0)
# Freeze all bn (affine) layers in resnet!!!
self.res5.apply(
lambda m: ResNet.freeze_params(m)
if isinstance(m, mynn.AffineChannel2d) else None)
self._init_weights()
Example #6
| Source File: ResNet.py From PMFNet with MIT License | 5 votes |
|
def _init_modules(self):
assert cfg.RESNETS.FREEZE_AT in [0, 2, 3, 4, 5]
assert cfg.RESNETS.FREEZE_AT <= self.convX
for i in range(1, cfg.RESNETS.FREEZE_AT + 1):
freeze_params(getattr(self, 'res%d' % i))
# Freeze all bn (affine) layers !!!
self.apply(lambda m: freeze_params(m) if isinstance(m, mynn.AffineChannel2d) else None)
Example #7
| Source File: ResNet.py From PMFNet with MIT License | 5 votes |
|
def _init_modules(self):
# Freeze all bn (affine) layers !!!
self.apply(lambda m: freeze_params(m) if isinstance(m, mynn.AffineChannel2d) else None)
Example #8
| Source File: ResNet.py From PMFNet with MIT License | 5 votes |
|
def basic_bn_shortcut(inplanes, outplanes, stride):
return nn.Sequential(
nn.Conv2d(inplanes,
outplanes,
kernel_size=1,
stride=stride,
bias=False),
mynn.AffineChannel2d(outplanes),
)
Example #9
| Source File: ResNet.py From PMFNet with MIT License | 5 votes |
|
def basic_bn_stem():
return nn.Sequential(OrderedDict([
('conv1', nn.Conv2d(3, 64, 7, stride=2, padding=3, bias=False)),
('bn1', mynn.AffineChannel2d(64)),
('relu', nn.ReLU(inplace=True)),
# ('maxpool', nn.MaxPool2d(kernel_size=3, stride=2, padding=0, ceil_mode=True))]))
('maxpool', nn.MaxPool2d(kernel_size=3, stride=2, padding=1))]))
Example #10
| Source File: mask_rcnn_heads.py From PMFNet with MIT License | 5 votes |
|
def __init__(self, dim_in, roi_xform_func, spatial_scale):
super().__init__()
self.dim_in = dim_in
self.roi_xform = roi_xform_func
self.spatial_scale = spatial_scale
self.dim_out = cfg.MRCNN.DIM_REDUCED
self.res5, dim_out = ResNet_roi_conv5_head_for_masks(dim_in)
self.upconv5 = nn.ConvTranspose2d(dim_out, self.dim_out, 2, 2, 0)
# Freeze all bn (affine) layers in resnet!!!
self.res5.apply(
lambda m: ResNet.freeze_params(m)
if isinstance(m, mynn.AffineChannel2d) else None)
self._init_weights()
Example #11
| Source File: ResNet.py From Large-Scale-VRD.pytorch with MIT License | 5 votes |
|
def _init_modules(self):
assert cfg.RESNETS.FREEZE_AT in [0, 2, 3, 4, 5]
assert cfg.RESNETS.FREEZE_AT <= self.convX
for i in range(1, cfg.RESNETS.FREEZE_AT + 1):
freeze_params(getattr(self, 'res%d' % i))
# Freeze all bn (affine) layers !!!
self.apply(lambda m: freeze_params(m) if isinstance(m, mynn.AffineChannel2d) else None)
Example #12
| Source File: ResNet.py From Large-Scale-VRD.pytorch with MIT License | 5 votes |
|
def _init_modules(self):
# Freeze all bn (affine) layers !!!
self.apply(lambda m: freeze_params(m) if isinstance(m, mynn.AffineChannel2d) else None)
Example #13
| Source File: ResNet.py From PANet with MIT License | 5 votes |
|
def _init_modules(self):
# Freeze all bn (affine) layers !!!
self.apply(lambda m: freeze_params(m) if isinstance(m, mynn.AffineChannel2d) else None)
Example #14
| Source File: ResNet.py From Large-Scale-VRD.pytorch with MIT License | 5 votes |
|
def basic_bn_stem():
return nn.Sequential(OrderedDict([
('conv1', nn.Conv2d(3, 64, 7, stride=2, padding=3, bias=False)),
('bn1', mynn.AffineChannel2d(64)),
('relu', nn.ReLU(inplace=True)),
# ('maxpool', nn.MaxPool2d(kernel_size=3, stride=2, padding=0, ceil_mode=True))]))
('maxpool', nn.MaxPool2d(kernel_size=3, stride=2, padding=1))]))
Example #15
| Source File: ResNet.py From detectron-self-train with MIT License | 5 votes |
|
def _init_modules(self):
assert cfg.RESNETS.FREEZE_AT in [0, 2, 3, 4, 5]
assert cfg.RESNETS.FREEZE_AT <= self.convX
for i in range(1, cfg.RESNETS.FREEZE_AT + 1):
freeze_params(getattr(self, 'res%d' % i))
# Freeze all bn (affine) layers !!!
self.apply(lambda m: freeze_params(m) if isinstance(m, mynn.AffineChannel2d) else None)
Example #16
| Source File: ResNet.py From detectron-self-train with MIT License | 5 votes |
|
def _init_modules(self):
# Freeze all bn (affine) layers !!!
self.apply(lambda m: freeze_params(m) if isinstance(m, mynn.AffineChannel2d) else None)
Example #17
| Source File: ResNet.py From detectron-self-train with MIT License | 5 votes |
|
def basic_bn_shortcut(inplanes, outplanes, stride):
return nn.Sequential(
nn.Conv2d(inplanes,
outplanes,
kernel_size=1,
stride=stride,
bias=False),
mynn.AffineChannel2d(outplanes),
)
Example #18
| Source File: ResNet.py From detectron-self-train with MIT License | 5 votes |
|
def basic_bn_stem():
return nn.Sequential(OrderedDict([
('conv1', nn.Conv2d(3, 64, 7, stride=2, padding=3, bias=False)),
('bn1', mynn.AffineChannel2d(64)),
('relu', nn.ReLU(inplace=True)),
# ('maxpool', nn.MaxPool2d(kernel_size=3, stride=2, padding=0, ceil_mode=True))]))
('maxpool', nn.MaxPool2d(kernel_size=3, stride=2, padding=1))]))
Example #19
| Source File: mask_rcnn_heads.py From detectron-self-train with MIT License | 5 votes |
|
def __init__(self, dim_in, roi_xform_func, spatial_scale):
super().__init__()
self.dim_in = dim_in
self.roi_xform = roi_xform_func
self.spatial_scale = spatial_scale
self.dim_out = cfg.MRCNN.DIM_REDUCED
self.res5, dim_out = ResNet_roi_conv5_head_for_masks(dim_in)
self.upconv5 = nn.ConvTranspose2d(dim_out, self.dim_out, 2, 2, 0)
# Freeze all bn (affine) layers in resnet!!!
self.res5.apply(
lambda m: ResNet.freeze_params(m)
if isinstance(m, mynn.AffineChannel2d) else None)
self._init_weights()
Example #20
| Source File: ResNet.py From DIoU-pytorch-detectron with GNU General Public License v3.0 | 5 votes |
|
def _init_modules(self):
assert cfg.RESNETS.FREEZE_AT in [0, 2, 3, 4, 5]
assert cfg.RESNETS.FREEZE_AT <= self.convX
for i in range(1, cfg.RESNETS.FREEZE_AT + 1):
freeze_params(getattr(self, 'res%d' % i))
# Freeze all bn (affine) layers !!!
self.apply(lambda m: freeze_params(m) if isinstance(m, mynn.AffineChannel2d) else None)
Example #21
| Source File: ResNet.py From DIoU-pytorch-detectron with GNU General Public License v3.0 | 5 votes |
|
def _init_modules(self):
# Freeze all bn (affine) layers !!!
self.apply(lambda m: freeze_params(m) if isinstance(m, mynn.AffineChannel2d) else None)
Example #22
| Source File: ResNet.py From DIoU-pytorch-detectron with GNU General Public License v3.0 | 5 votes |
|
def basic_bn_shortcut(inplanes, outplanes, stride):
return nn.Sequential(
nn.Conv2d(inplanes,
outplanes,
kernel_size=1,
stride=stride,
bias=False),
mynn.AffineChannel2d(outplanes),
)
Example #23
| Source File: ResNet.py From DIoU-pytorch-detectron with GNU General Public License v3.0 | 5 votes |
|
def basic_bn_stem():
return nn.Sequential(OrderedDict([
('conv1', nn.Conv2d(3, 64, 7, stride=2, padding=3, bias=False)),
('bn1', mynn.AffineChannel2d(64)),
('relu', nn.ReLU(inplace=True)),
# ('maxpool', nn.MaxPool2d(kernel_size=3, stride=2, padding=0, ceil_mode=True))]))
('maxpool', nn.MaxPool2d(kernel_size=3, stride=2, padding=1))]))
Example #24
| Source File: mask_rcnn_heads.py From DIoU-pytorch-detectron with GNU General Public License v3.0 | 5 votes |
|
def __init__(self, dim_in, roi_xform_func, spatial_scale):
super().__init__()
self.dim_in = dim_in
self.roi_xform = roi_xform_func
self.spatial_scale = spatial_scale
self.dim_out = cfg.MRCNN.DIM_REDUCED
self.res5, dim_out = ResNet_roi_conv5_head_for_masks(dim_in)
self.upconv5 = nn.ConvTranspose2d(dim_out, self.dim_out, 2, 2, 0)
# Freeze all bn (affine) layers in resnet!!!
self.res5.apply(
lambda m: ResNet.freeze_params(m)
if isinstance(m, mynn.AffineChannel2d) else None)
self._init_weights()
Example #25
| Source File: ResNet.py From Detectron.pytorch with MIT License | 5 votes |
|
def _init_modules(self):
assert cfg.RESNETS.FREEZE_AT in [0, 2, 3, 4, 5]
assert cfg.RESNETS.FREEZE_AT <= self.convX
for i in range(1, cfg.RESNETS.FREEZE_AT + 1):
freeze_params(getattr(self, 'res%d' % i))
# Freeze all bn (affine) layers !!!
self.apply(lambda m: freeze_params(m) if isinstance(m, mynn.AffineChannel2d) else None)
Example #26
| Source File: ResNet.py From Detectron.pytorch with MIT License | 5 votes |
|
def _init_modules(self):
assert cfg.RESNETS.FREEZE_AT in [0, 2, 3, 4, 5]
assert cfg.RESNETS.FREEZE_AT <= self.convX
for i in range(1, cfg.RESNETS.FREEZE_AT + 1):
freeze_params(getattr(self, 'res%d' % i))
# Freeze all bn (affine) layers !!!
self.apply(lambda m: freeze_params(m) if isinstance(m, mynn.AffineChannel2d) else None)
Example #27
| Source File: ResNet.py From Detectron.pytorch with MIT License | 5 votes |
|
def _init_modules(self):
# Freeze all bn (affine) layers !!!
self.apply(lambda m: freeze_params(m) if isinstance(m, mynn.AffineChannel2d) else None)
Example #28
| Source File: ResNet.py From Detectron.pytorch with MIT License | 5 votes |
|
def basic_bn_shortcut(inplanes, outplanes, stride):
return nn.Sequential(
nn.Conv2d(inplanes,
outplanes,
kernel_size=1,
stride=stride,
bias=False),
mynn.AffineChannel2d(outplanes),
)
Example #29
| Source File: ResNet.py From Detectron.pytorch with MIT License | 5 votes |
|
def basic_bn_stem():
return nn.Sequential(OrderedDict([
('conv1', nn.Conv2d(3, 64, 7, stride=2, padding=3, bias=False)),
('bn1', mynn.AffineChannel2d(64)),
('relu', nn.ReLU(inplace=True)),
# ('maxpool', nn.MaxPool2d(kernel_size=3, stride=2, padding=0, ceil_mode=True))]))
('maxpool', nn.MaxPool2d(kernel_size=3, stride=2, padding=1))]))
Example #30
| Source File: mask_rcnn_heads.py From Detectron.pytorch with MIT License | 5 votes |
|
def __init__(self, dim_in, roi_xform_func, spatial_scale):
super().__init__()
self.dim_in = dim_in
self.roi_xform = roi_xform_func
self.spatial_scale = spatial_scale
self.dim_out = cfg.MRCNN.DIM_REDUCED
self.res5, dim_out = ResNet_roi_conv5_head_for_masks(dim_in)
self.upconv5 = nn.ConvTranspose2d(dim_out, self.dim_out, 2, 2, 0)
# Freeze all bn (affine) layers in resnet!!!
self.res5.apply(
lambda m: ResNet.freeze_params(m)
if isinstance(m, mynn.AffineChannel2d) else None)
self._init_weights()
