Skip to content

GitLab

"git@developer.sourcefind.cn:OpenDAS/ollama.git" did not exist on "074bac844714b3a011d1cf6ebc40c169b10d1aef"
Commit 03c38bdc authored by yhcao6's avatar yhcao6
Browse files

separate dcn and dpool cpp, restruct some code

parent 66489d6b
Expand all Hide whitespace changes
Inline Side-by-side
Showing with 822 additions and 493 deletions
mmdet/ops/dcn/__init__.py
View file @ 03c38bdc
...@@ -2,11 +2,12 @@ from .functions.deform_conv import deform_conv, modulated_deform_conv ...@@ -2,11 +2,12 @@ from .functions.deform_conv import deform_conv, modulated_deform_conv
from .functions.deform_pool import deform_roi_pooling from .functions.deform_pool import deform_roi_pooling
from .modules.deform_conv import (DeformConv, ModulatedDeformConv, from .modules.deform_conv import (DeformConv, ModulatedDeformConv,
ModulatedDeformConvPack) ModulatedDeformConvPack)
from .modules.deform_pool import (DeformRoIPooling, from .modules.deform_pool import (DeformRoIPooling, DeformRoIPoolingPack,
ModulatedDeformRoIPoolingPack) ModulatedDeformRoIPoolingPack)
__all__ = [ __all__ = [
'DeformConv', 'DeformRoIPooling', 'ModulatedDeformRoIPoolingPack', 'DeformConv', 'DeformRoIPooling', 'DeformRoIPoolingPack',
'ModulatedDeformConv', 'ModulatedDeformConvPack', 'deform_conv', 'ModulatedDeformRoIPoolingPack', 'ModulatedDeformConv',
'ModulatedDeformConvPack', 'deform_conv',
'modulated_deform_conv', 'deform_roi_pooling' 'modulated_deform_conv', 'deform_roi_pooling'
] ]
mmdet/ops/dcn/functions/deform_conv.py
View file @ 03c38bdc
...@@ -3,7 +3,6 @@ from torch.autograd import Function ...@@ -3,7 +3,6 @@ from torch.autograd import Function
from torch.nn.modules.utils import _pair from torch.nn.modules.utils import _pair
from .. import deform_conv_cuda from .. import deform_conv_cuda
from .. import modulated_dcn_cuda as _backend
class DeformConvFunction(Function): class DeformConvFunction(Function):
...@@ -124,7 +123,7 @@ class ModulatedDeformConvFunction(Function): ...@@ -124,7 +123,7 @@ class ModulatedDeformConvFunction(Function):
output = input.new( output = input.new(
*ModulatedDeformConvFunction._infer_shape(ctx, input, weight)) *ModulatedDeformConvFunction._infer_shape(ctx, input, weight))
ctx._bufs = [input.new(), input.new()] ctx._bufs = [input.new(), input.new()]
_backend.modulated_deform_conv_cuda_forward( deform_conv_cuda.modulated_deform_conv_cuda_forward(
input, weight, bias, ctx._bufs[0], offset, mask, output, input, weight, bias, ctx._bufs[0], offset, mask, output,
ctx._bufs[1], weight.shape[2], weight.shape[3], ctx.stride, ctx._bufs[1], weight.shape[2], weight.shape[3], ctx.stride,
ctx.stride, ctx.padding, ctx.padding, ctx.dilation, ctx.dilation, ctx.stride, ctx.padding, ctx.padding, ctx.dilation, ctx.dilation,
...@@ -141,7 +140,7 @@ class ModulatedDeformConvFunction(Function): ...@@ -141,7 +140,7 @@ class ModulatedDeformConvFunction(Function):
grad_mask = torch.zeros_like(mask) grad_mask = torch.zeros_like(mask)
grad_weight = torch.zeros_like(weight) grad_weight = torch.zeros_like(weight)
grad_bias = torch.zeros_like(bias) grad_bias = torch.zeros_like(bias)
_backend.modulated_deform_conv_cuda_backward( deform_conv_cuda.modulated_deform_conv_cuda_backward(
input, weight, bias, ctx._bufs[0], offset, mask, ctx._bufs[1], input, weight, bias, ctx._bufs[0], offset, mask, ctx._bufs[1],
grad_input, grad_weight, grad_bias, grad_offset, grad_mask, grad_input, grad_weight, grad_bias, grad_offset, grad_mask,
grad_output, weight.shape[2], weight.shape[3], ctx.stride, grad_output, weight.shape[2], weight.shape[3], ctx.stride,
......
mmdet/ops/dcn/functions/deform_pool.py
View file @ 03c38bdc
import torch import torch
from torch.autograd import Function from torch.autograd import Function
from .. import modulated_dcn_cuda as _backend from .. import deform_pool_cuda
class DeformRoIPoolingFunction(Function): class DeformRoIPoolingFunction(Function):
...@@ -36,7 +36,7 @@ class DeformRoIPoolingFunction(Function): ...@@ -36,7 +36,7 @@ class DeformRoIPoolingFunction(Function):
*DeformRoIPoolingFunction._infer_shape(ctx, data, rois)) *DeformRoIPoolingFunction._infer_shape(ctx, data, rois))
output_count = data.new( output_count = data.new(
*DeformRoIPoolingFunction._infer_shape(ctx, data, rois)) *DeformRoIPoolingFunction._infer_shape(ctx, data, rois))
_backend.deform_psroi_pooling_cuda_forward( deform_pool_cuda.deform_psroi_pooling_cuda_forward(
data, rois, offset, output, output_count, ctx.no_trans, data, rois, offset, output, output_count, ctx.no_trans,
ctx.spatial_scale, ctx.output_dim, ctx.group_size, ctx.pooled_size, ctx.spatial_scale, ctx.output_dim, ctx.group_size, ctx.pooled_size,
ctx.part_size, ctx.sample_per_part, ctx.trans_std) ctx.part_size, ctx.sample_per_part, ctx.trans_std)
...@@ -63,7 +63,7 @@ class DeformRoIPoolingFunction(Function): ...@@ -63,7 +63,7 @@ class DeformRoIPoolingFunction(Function):
grad_input = torch.zeros_like(data) grad_input = torch.zeros_like(data)
grad_offset = torch.zeros_like(offset) grad_offset = torch.zeros_like(offset)
_backend.deform_psroi_pooling_cuda_backward( deform_pool_cuda.deform_psroi_pooling_cuda_backward(
grad_output, data, rois, offset, output_count, grad_input, grad_output, data, rois, offset, output_count, grad_input,
grad_offset, ctx.no_trans, ctx.spatial_scale, ctx.output_dim, grad_offset, ctx.no_trans, ctx.spatial_scale, ctx.output_dim,
ctx.group_size, ctx.pooled_size, ctx.part_size, ctx.group_size, ctx.pooled_size, ctx.part_size,
......
mmdet/ops/dcn/modules/deform_conv.py
View file @ 03c38bdc
...@@ -16,7 +16,7 @@ class DeformConv(nn.Module): ...@@ -16,7 +16,7 @@ class DeformConv(nn.Module):
stride=1, stride=1,
padding=0, padding=0,
dilation=1, dilation=1,
num_deformable_groups=1): deformable_groups=1):
super(DeformConv, self).__init__() super(DeformConv, self).__init__()
self.in_channels = in_channels self.in_channels = in_channels
self.out_channels = out_channels self.out_channels = out_channels
...@@ -24,7 +24,7 @@ class DeformConv(nn.Module): ...@@ -24,7 +24,7 @@ class DeformConv(nn.Module):
self.stride = _pair(stride) self.stride = _pair(stride)
self.padding = _pair(padding) self.padding = _pair(padding)
self.dilation = _pair(dilation) self.dilation = _pair(dilation)
self.num_deformable_groups = num_deformable_groups self.deformable_groups = deformable_groups
self.weight = nn.Parameter( self.weight = nn.Parameter(
torch.Tensor(out_channels, in_channels, *self.kernel_size)) torch.Tensor(out_channels, in_channels, *self.kernel_size))
...@@ -41,7 +41,7 @@ class DeformConv(nn.Module): ...@@ -41,7 +41,7 @@ class DeformConv(nn.Module):
def forward(self, input, offset): def forward(self, input, offset):
return deform_conv(input, offset, self.weight, self.stride, return deform_conv(input, offset, self.weight, self.stride,
self.padding, self.dilation, self.padding, self.dilation,
self.num_deformable_groups) self.deformable_groups)
class ModulatedDeformConv(nn.Module): class ModulatedDeformConv(nn.Module):
...@@ -54,7 +54,7 @@ class ModulatedDeformConv(nn.Module): ...@@ -54,7 +54,7 @@ class ModulatedDeformConv(nn.Module):
padding, padding,
dilation=1, dilation=1,
deformable_groups=1, deformable_groups=1,
no_bias=True): bias=False):
super(ModulatedDeformConv, self).__init__() super(ModulatedDeformConv, self).__init__()
self.in_channels = in_channels self.in_channels = in_channels
self.out_channels = out_channels self.out_channels = out_channels
...@@ -63,13 +63,12 @@ class ModulatedDeformConv(nn.Module): ...@@ -63,13 +63,12 @@ class ModulatedDeformConv(nn.Module):
self.padding = padding self.padding = padding
self.dilation = dilation self.dilation = dilation
self.deformable_groups = deformable_groups self.deformable_groups = deformable_groups
self.no_bias = no_bias
self.weight = nn.Parameter( self.weight = nn.Parameter(
torch.Tensor(out_channels, in_channels, *self.kernel_size)) torch.Tensor(out_channels, in_channels, *self.kernel_size))
self.bias = nn.Parameter(torch.zeros(out_channels)) self.bias = nn.Parameter(torch.zeros(out_channels))
self.reset_parameters() self.reset_parameters()
if self.no_bias: if not bias:
self.bias.requires_grad = False self.bias.requires_grad = False
def reset_parameters(self): def reset_parameters(self):
...@@ -96,10 +95,10 @@ class ModulatedDeformConvPack(ModulatedDeformConv): ...@@ -96,10 +95,10 @@ class ModulatedDeformConvPack(ModulatedDeformConv):
padding, padding,
dilation=1, dilation=1,
deformable_groups=1, deformable_groups=1,
no_bias=False): bias=True):
super(ModulatedDeformConvPack, super(ModulatedDeformConvPack,
self).__init__(in_channels, out_channels, kernel_size, stride, self).__init__(in_channels, out_channels, kernel_size, stride,
padding, dilation, deformable_groups, no_bias) padding, dilation, deformable_groups, bias)
self.conv_offset_mask = nn.Conv2d( self.conv_offset_mask = nn.Conv2d(
self.in_channels, self.in_channels,
......
mmdet/ops/dcn/modules/deform_pool.py
View file @ 03c38bdc
...@@ -7,7 +7,7 @@ class DeformRoIPooling(nn.Module): ...@@ -7,7 +7,7 @@ class DeformRoIPooling(nn.Module):
def __init__(self, def __init__(self,
spatial_scale, spatial_scale,
pooled_size, out_size,
output_dim, output_dim,
no_trans, no_trans,
group_size=1, group_size=1,
...@@ -16,12 +16,11 @@ class DeformRoIPooling(nn.Module): ...@@ -16,12 +16,11 @@ class DeformRoIPooling(nn.Module):
trans_std=.0): trans_std=.0):
super(DeformRoIPooling, self).__init__() super(DeformRoIPooling, self).__init__()
self.spatial_scale = spatial_scale self.spatial_scale = spatial_scale
self.pooled_size = pooled_size self.out_size = out_size
self.out_size = pooled_size
self.output_dim = output_dim self.output_dim = output_dim
self.no_trans = no_trans self.no_trans = no_trans
self.group_size = group_size self.group_size = group_size
self.part_size = pooled_size if part_size is None else part_size self.part_size = out_size if part_size is None else part_size
self.sample_per_part = sample_per_part self.sample_per_part = sample_per_part
self.trans_std = trans_std self.trans_std = trans_std
...@@ -29,7 +28,7 @@ class DeformRoIPooling(nn.Module): ...@@ -29,7 +28,7 @@ class DeformRoIPooling(nn.Module):
if self.no_trans: if self.no_trans:
offset = data.new() offset = data.new()
return deform_roi_pooling( return deform_roi_pooling(
data, rois, offset, self.spatial_scale, self.pooled_size, data, rois, offset, self.spatial_scale, self.out_size,
self.output_dim, self.no_trans, self.group_size, self.part_size, self.output_dim, self.no_trans, self.group_size, self.part_size,
self.sample_per_part, self.trans_std) self.sample_per_part, self.trans_std)
...@@ -38,7 +37,7 @@ class ModulatedDeformRoIPoolingPack(DeformRoIPooling): ...@@ -38,7 +37,7 @@ class ModulatedDeformRoIPoolingPack(DeformRoIPooling):
def __init__(self, def __init__(self,
spatial_scale, spatial_scale,
pooled_size, out_size,
output_dim, output_dim,
no_trans, no_trans,
group_size=1, group_size=1,
...@@ -47,7 +46,7 @@ class ModulatedDeformRoIPoolingPack(DeformRoIPooling): ...@@ -47,7 +46,7 @@ class ModulatedDeformRoIPoolingPack(DeformRoIPooling):
trans_std=.0, trans_std=.0,
deform_fc_dim=1024): deform_fc_dim=1024):
super(ModulatedDeformRoIPoolingPack, self).__init__( super(ModulatedDeformRoIPoolingPack, self).__init__(
spatial_scale, pooled_size, output_dim, no_trans, group_size, spatial_scale, out_size, output_dim, no_trans, group_size,
part_size, sample_per_part, trans_std) part_size, sample_per_part, trans_std)
self.deform_fc_dim = deform_fc_dim self.deform_fc_dim = deform_fc_dim
...@@ -55,20 +54,20 @@ class ModulatedDeformRoIPoolingPack(DeformRoIPooling): ...@@ -55,20 +54,20 @@ class ModulatedDeformRoIPoolingPack(DeformRoIPooling):
if not no_trans: if not no_trans:
self.offset_fc = nn.Sequential( self.offset_fc = nn.Sequential(
nn.Linear( nn.Linear(
self.pooled_size * self.pooled_size * self.output_dim, self.out_size * self.out_size * self.output_dim,
self.deform_fc_dim), nn.ReLU(inplace=True), self.deform_fc_dim), nn.ReLU(inplace=True),
nn.Linear(self.deform_fc_dim, self.deform_fc_dim), nn.Linear(self.deform_fc_dim, self.deform_fc_dim),
nn.ReLU(inplace=True), nn.ReLU(inplace=True),
nn.Linear(self.deform_fc_dim, nn.Linear(self.deform_fc_dim,
self.pooled_size * self.pooled_size * 2)) self.out_size * self.out_size * 2))
self.offset_fc[4].weight.data.zero_() self.offset_fc[4].weight.data.zero_()
self.offset_fc[4].bias.data.zero_() self.offset_fc[4].bias.data.zero_()
self.mask_fc = nn.Sequential( self.mask_fc = nn.Sequential(
nn.Linear( nn.Linear(
self.pooled_size * self.pooled_size * self.output_dim, self.out_size * self.out_size * self.output_dim,
self.deform_fc_dim), nn.ReLU(inplace=True), self.deform_fc_dim), nn.ReLU(inplace=True),
nn.Linear(self.deform_fc_dim, nn.Linear(self.deform_fc_dim,
self.pooled_size * self.pooled_size * 1), self.out_size * self.out_size * 1),
nn.Sigmoid()) nn.Sigmoid())
self.mask_fc[2].weight.data.zero_() self.mask_fc[2].weight.data.zero_()
self.mask_fc[2].bias.data.zero_() self.mask_fc[2].bias.data.zero_()
...@@ -80,19 +79,72 @@ class ModulatedDeformRoIPoolingPack(DeformRoIPooling): ...@@ -80,19 +79,72 @@ class ModulatedDeformRoIPoolingPack(DeformRoIPooling):
n = rois.shape[0] n = rois.shape[0]
offset = data.new() offset = data.new()
x = deform_roi_pooling(data, rois, offset, self.spatial_scale, x = deform_roi_pooling(data, rois, offset, self.spatial_scale,
self.pooled_size, self.output_dim, True, self.out_size, self.output_dim, True,
self.group_size, self.part_size, self.group_size, self.part_size,
self.sample_per_part, self.trans_std) self.sample_per_part, self.trans_std)
offset = self.offset_fc(x.view(n, -1)) offset = self.offset_fc(x.view(n, -1))
offset = offset.view(n, 2, self.pooled_size, self.pooled_size) offset = offset.view(n, 2, self.out_size, self.out_size)
mask = self.mask_fc(x.view(n, -1)) mask = self.mask_fc(x.view(n, -1))
mask = mask.view(n, 1, self.pooled_size, self.pooled_size) mask = mask.view(n, 1, self.out_size, self.out_size)
feat = deform_roi_pooling( feat = deform_roi_pooling(
data, rois, offset, self.spatial_scale, self.pooled_size, data, rois, offset, self.spatial_scale, self.out_size,
self.output_dim, self.no_trans, self.group_size, self.output_dim, self.no_trans, self.group_size,
self.part_size, self.sample_per_part, self.trans_std) * mask self.part_size, self.sample_per_part, self.trans_std) * mask
return feat return feat
return deform_roi_pooling( return deform_roi_pooling(
data, rois, offset, self.spatial_scale, self.pooled_size, data, rois, offset, self.spatial_scale, self.out_size,
self.output_dim, self.no_trans, self.group_size, self.part_size,
self.sample_per_part, self.trans_std)
class DeformRoIPoolingPack(DeformRoIPooling):
def __init__(self,
spatial_scale,
out_size,
output_dim,
no_trans,
group_size=1,
part_size=None,
sample_per_part=4,
trans_std=.0,
deform_fc_dim=1024):
super(DeformRoIPoolingPack, self).__init__(
spatial_scale, out_size, output_dim, no_trans, group_size,
part_size, sample_per_part, trans_std)
self.deform_fc_dim = deform_fc_dim
if not no_trans:
self.offset_fc = nn.Sequential(
nn.Linear(
self.out_size * self.out_size * self.output_dim,
self.deform_fc_dim), nn.ReLU(inplace=True),
nn.Linear(self.deform_fc_dim, self.deform_fc_dim),
nn.ReLU(inplace=True),
nn.Linear(self.deform_fc_dim,
self.out_size * self.out_size * 2))
self.offset_fc[4].weight.data.zero_()
self.offset_fc[4].bias.data.zero_()
def forward(self, data, rois):
if self.no_trans:
offset = data.new()
else:
n = rois.shape[0]
offset = data.new()
x = deform_roi_pooling(data, rois, offset, self.spatial_scale,
self.out_size, self.output_dim, True,
self.group_size, self.part_size,
self.sample_per_part, self.trans_std)
offset = self.offset_fc(x.view(n, -1))
offset = offset.view(n, 2, self.out_size, self.out_size)
feat = deform_roi_pooling(
data, rois, offset, self.spatial_scale, self.out_size,
self.output_dim, self.no_trans, self.group_size,
self.part_size, self.sample_per_part, self.trans_std)
return feat
return deform_roi_pooling(
data, rois, offset, self.spatial_scale, self.out_size,
self.output_dim, self.no_trans, self.group_size, self.part_size, self.output_dim, self.no_trans, self.group_size, self.part_size,
self.sample_per_part, self.trans_std) self.sample_per_part, self.trans_std)
mmdet/ops/dcn/setup.py
View file @ 03c38bdc
...@@ -2,11 +2,14 @@ from setuptools import setup ...@@ -2,11 +2,14 @@ from setuptools import setup
from torch.utils.cpp_extension import BuildExtension, CUDAExtension from torch.utils.cpp_extension import BuildExtension, CUDAExtension
setup( setup(
name='deform_conv_cuda', name='deform_conv',
ext_modules=[ ext_modules=[
CUDAExtension('deform_conv_cuda', [ CUDAExtension('deform_conv_cuda', [
'src/deform_conv_cuda.cpp', 'src/deform_conv_cuda.cpp',
'src/deform_conv_cuda_kernel.cu', 'src/deform_conv_cuda_kernel.cu',
]), ]),
CUDAExtension('deform_pool_cuda', [
'src/deform_pool_cuda.cpp', 'src/deform_pool_cuda_kernel.cu'
]),
], ],
cmdclass={'build_ext': BuildExtension}) cmdclass={'build_ext': BuildExtension})
mmdet/ops/dcn/setup_modulated.py deleted 100644 → 0
View file @ 66489d6b
from setuptools import setup
from torch.utils.cpp_extension import BuildExtension, CUDAExtension
setup(
name='modulated_dcn_cuda',
ext_modules=[
CUDAExtension('modulated_dcn_cuda', [
'src/modulated_dcn_cuda.cpp',
'src/modulated_deform_im2col_cuda.cu',
'src/deform_psroi_pooling_cuda.cu'
]),
],
cmdclass={'build_ext': BuildExtension})
mmdet/ops/dcn/src/deform_conv_cuda.cpp
View file @ 03c38bdc
...@@ -33,6 +33,32 @@ void deformable_col2im_coord(const at::Tensor data_col, ...@@ -33,6 +33,32 @@ void deformable_col2im_coord(const at::Tensor data_col,
const int dilation_w, const int parallel_imgs, const int dilation_w, const int parallel_imgs,
const int deformable_group, at::Tensor grad_offset); const int deformable_group, at::Tensor grad_offset);
void modulated_deformable_im2col_cuda(const at::Tensor data_im, const at::Tensor data_offset,
const at::Tensor data_mask, const int batch_size, const int channels,
const int height_im, const int width_im, const int height_col,
const int width_col, const int kernel_h, const int kenerl_w,
const int pad_h, const int pad_w, const int stride_h, const int stride_w,
const int dilation_h, const int dilation_w,
const int deformable_group, at::Tensor data_col);
void modulated_deformable_col2im_cuda(const at::Tensor data_col, const at::Tensor data_offset,
const at::Tensor data_mask, const int batch_size, const int channels,
const int height_im, const int width_im, const int height_col,
const int width_col, const int kernel_h, const int kenerl_w,
const int pad_h, const int pad_w, const int stride_h, const int stride_w,
const int dilation_h, const int dilation_w,
const int deformable_group, at::Tensor grad_im);
void modulated_deformable_col2im_coord_cuda(const at::Tensor data_col, const at::Tensor data_im,
const at::Tensor data_offset, const at::Tensor data_mask,
const int batch_size, const int channels, const int height_im,
const int width_im, const int height_col, const int width_col,
const int kernel_h, const int kenerl_w, const int pad_h,
const int pad_w, const int stride_h, const int stride_w,
const int dilation_h, const int dilation_w,
const int deformable_group, at::Tensor grad_offset,
at::Tensor grad_mask);
void shape_check(at::Tensor input, at::Tensor offset, void shape_check(at::Tensor input, at::Tensor offset,
at::Tensor *gradOutput, at::Tensor weight, int kH, int kW, at::Tensor *gradOutput, at::Tensor weight, int kH, int kW,
int dH, int dW, int padH, int padW, int dilationH, int dH, int dW, int padH, int padW, int dilationH,
...@@ -256,16 +282,6 @@ int deform_conv_backward_input_cuda( ...@@ -256,16 +282,6 @@ int deform_conv_backward_input_cuda(
{batchSize / im2col_step, im2col_step, nOutputPlane, outputHeight, outputWidth}); {batchSize / im2col_step, im2col_step, nOutputPlane, outputHeight, outputWidth});
gradOutput.transpose_(1, 2); gradOutput.transpose_(1, 2);
at::Tensor gradOutputBuffer = at::zeros_like(gradOutput);
gradOutputBuffer = gradOutputBuffer.view(
{batchSize / im2col_step, nOutputPlane, im2col_step, outputHeight, outputWidth});
gradOutputBuffer.copy_(gradOutput);
gradOutputBuffer = gradOutputBuffer.view(
{batchSize / im2col_step, nOutputPlane, im2col_step * outputHeight, outputWidth});
gradOutput.transpose_(1, 2);
gradOutput = gradOutput.view({batchSize, nOutputPlane, outputHeight, outputWidth});
gradInput = gradInput.view( gradInput = gradInput.view(
{batchSize / im2col_step, im2col_step, nInputPlane, inputHeight, inputWidth}); {batchSize / im2col_step, im2col_step, nInputPlane, inputHeight, inputWidth});
input = input.view({batchSize / im2col_step, im2col_step, nInputPlane, inputHeight, inputWidth}); input = input.view({batchSize / im2col_step, im2col_step, nInputPlane, inputHeight, inputWidth});
...@@ -276,7 +292,7 @@ int deform_conv_backward_input_cuda( ...@@ -276,7 +292,7 @@ int deform_conv_backward_input_cuda(
for (int elt = 0; elt < batchSize / im2col_step; elt++) for (int elt = 0; elt < batchSize / im2col_step; elt++)
{ {
columns = columns.addmm_(weight.flatten(1).transpose(0, 1), gradOutputBuffer[elt].flatten(1), 0.0f, 1.0f); columns = columns.addmm_(weight.flatten(1).transpose(0, 1), gradOutput[elt].flatten(1), 0.0f, 1.0f);
deformable_col2im_coord( deformable_col2im_coord(
columns, input[elt], offset[elt], columns, input[elt], offset[elt],
...@@ -289,6 +305,9 @@ int deform_conv_backward_input_cuda( ...@@ -289,6 +305,9 @@ int deform_conv_backward_input_cuda(
deformable_group, gradInput[elt]); deformable_group, gradInput[elt]);
} }
gradOutput.transpose_(1, 2);
gradOutput = gradOutput.view({batchSize, nOutputPlane, outputHeight, outputWidth});
gradInput = gradInput.view({batchSize, nInputPlane, inputHeight, inputWidth}); gradInput = gradInput.view({batchSize, nInputPlane, inputHeight, inputWidth});
input = input.view({batchSize, nInputPlane, inputHeight, inputWidth}); input = input.view({batchSize, nInputPlane, inputHeight, inputWidth});
gradOffset = gradOffset.view({batchSize, deformable_group * 2 * kH * kW, outputHeight, outputWidth}); gradOffset = gradOffset.view({batchSize, deformable_group * 2 * kH * kW, outputHeight, outputWidth});
...@@ -394,6 +413,148 @@ int deform_conv_backward_parameters_cuda( ...@@ -394,6 +413,148 @@ int deform_conv_backward_parameters_cuda(
return 1; return 1;
} }
void modulated_deform_conv_cuda_forward(at::Tensor input, at::Tensor weight,
at::Tensor bias, at::Tensor ones,
at::Tensor offset, at::Tensor mask,
at::Tensor output, at::Tensor columns,
int kernel_h, int kernel_w,
const int stride_h, const int stride_w,
const int pad_h, const int pad_w,
const int dilation_h, const int dilation_w,
const int deformable_group)
{
AT_CHECK(input.is_contiguous(), "input tensor has to be contiguous");
AT_CHECK(weight.is_contiguous(), "weight tensor has to be contiguous");
const int batch = input.size(0);
const int channels = input.size(1);
const int height = input.size(2);
const int width = input.size(3);
const int channels_out = weight.size(0);
const int channels_kernel = weight.size(1);
const int kernel_h_ = weight.size(2);
const int kernel_w_ = weight.size(3);
if (kernel_h_ != kernel_h || kernel_w_ != kernel_w)
AT_ERROR("Input shape and kernel shape wont match: (%d x %d vs %d x %d).",
kernel_h_, kernel_w, kernel_h_, kernel_w_);
if (channels != channels_kernel)
AT_ERROR("Input shape and kernel channels wont match: (%d vs %d).",
channels, channels_kernel);
const int height_out = (height + 2 * pad_h - (dilation_h * (kernel_h - 1) + 1)) / stride_h + 1;
const int width_out = (width + 2 * pad_w - (dilation_w * (kernel_w - 1) + 1)) / stride_w + 1;
if (ones.ndimension() != 2 ||
ones.size(0) * ones.size(1) < height_out * width_out)
{
// Resize plane and fill with ones...
ones = at::ones({height_out, width_out}, input.type());
}
// resize output
output = output.view({batch, channels_out, height_out, width_out});
// resize temporary columns
columns = at::zeros({channels * kernel_h * kernel_w, 1 * height_out * width_out}, input.type());
for (int b = 0; b < batch; b++)
{
// Do Bias first:
// M,N,K are dims of matrix A and B
// (see http://docs.nvidia.com/cuda/cublas/#cublas-lt-t-gt-gemm)
// (N x 1) (1 x M)
output[b] = output[b].flatten(1).addmm_(bias.view({-1, 1}), ones.view({1, -1}), 0.0f, 1.0f).view_as(output[b]);
modulated_deformable_im2col_cuda(input[b], offset[b], mask[b],
1, channels, height, width,
height_out, width_out, kernel_h, kernel_w,
pad_h, pad_w, stride_h, stride_w, dilation_h, dilation_w,
deformable_group, columns);
//(k * m) x (m * n)
// Y = WC
output[b] = output[b].flatten(1).addmm_(weight.flatten(1), columns).view_as(output[b]);
}
}
void modulated_deform_conv_cuda_backward(at::Tensor input, at::Tensor weight,
at::Tensor bias, at::Tensor ones,
at::Tensor offset, at::Tensor mask,
at::Tensor columns,
at::Tensor grad_input, at::Tensor grad_weight,
at::Tensor grad_bias, at::Tensor grad_offset,
at::Tensor grad_mask, at::Tensor grad_output,
int kernel_h, int kernel_w,
int stride_h, int stride_w,
int pad_h, int pad_w,
int dilation_h, int dilation_w,
int deformable_group)
{
AT_CHECK(input.is_contiguous(), "input tensor has to be contiguous");
AT_CHECK(weight.is_contiguous(), "weight tensor has to be contiguous");
const int batch = input.size(0);
const int channels = input.size(1);
const int height = input.size(2);
const int width = input.size(3);
const int channels_kernel = weight.size(1);
const int kernel_h_ = weight.size(2);
const int kernel_w_ = weight.size(3);
if (kernel_h_ != kernel_h || kernel_w_ != kernel_w)
AT_ERROR("Input shape and kernel shape wont match: (%d x %d vs %d x %d).",
kernel_h_, kernel_w, kernel_h_, kernel_w_);
if (channels != channels_kernel)
AT_ERROR("Input shape and kernel channels wont match: (%d vs %d).",
channels, channels_kernel);
const int height_out = (height + 2 * pad_h - (dilation_h * (kernel_h - 1) + 1)) / stride_h + 1;
const int width_out = (width + 2 * pad_w - (dilation_w * (kernel_w - 1) + 1)) / stride_w + 1;
if (ones.ndimension() != 2 ||
ones.size(0) * ones.size(1) < height_out * width_out)
{
// Resize plane and fill with ones...
ones = at::ones({height_out, width_out}, input.type());
}
grad_input = grad_input.view({batch, channels, height, width});
columns = at::zeros({channels * kernel_h * kernel_w, height_out * width_out}, input.type());
for (int b = 0; b < batch; b++)
{
columns.addmm_(weight.flatten(1).transpose(0, 1), grad_output[b].flatten(1), 0.0f, 1.0f);
// gradient w.r.t. input coordinate data
modulated_deformable_col2im_coord_cuda(columns, input[b], offset[b], mask[b],
1, channels, height, width,
height_out, width_out, kernel_h, kernel_w,
pad_h, pad_w, stride_h, stride_w,
dilation_h, dilation_w, deformable_group,
grad_offset[b], grad_mask[b]);
// gradient w.r.t. input data
modulated_deformable_col2im_cuda(columns, offset[b], mask[b],
1, channels, height, width,
height_out, width_out, kernel_h, kernel_w,
pad_h, pad_w, stride_h, stride_w,
dilation_h, dilation_w, deformable_group,
grad_input[b]);
// gradient w.r.t. weight, dWeight should accumulate across the batch and group
modulated_deformable_im2col_cuda(input[b], offset[b], mask[b],
1, channels, height, width,
height_out, width_out, kernel_h, kernel_w,
pad_h, pad_w, stride_h, stride_w,
dilation_h, dilation_w, deformable_group,
columns);
grad_weight = grad_weight.flatten(1).addmm_(grad_output[b].flatten(1), columns.transpose(0, 1)).view_as(grad_weight);
grad_bias = grad_bias.view({-1, 1}).addmm_(grad_output[b].flatten(1), ones.view({-1, 1})).view(-1);
}
}
PYBIND11_MODULE(TORCH_EXTENSION_NAME, m) PYBIND11_MODULE(TORCH_EXTENSION_NAME, m)
{ {
m.def("deform_conv_forward_cuda", &deform_conv_forward_cuda, "deform forward (CUDA)"); m.def("deform_conv_forward_cuda", &deform_conv_forward_cuda, "deform forward (CUDA)");
...@@ -401,4 +562,8 @@ PYBIND11_MODULE(TORCH_EXTENSION_NAME, m) ...@@ -401,4 +562,8 @@ PYBIND11_MODULE(TORCH_EXTENSION_NAME, m)
"deform_conv_backward_input (CUDA)"); "deform_conv_backward_input (CUDA)");
m.def("deform_conv_backward_parameters_cuda", &deform_conv_backward_parameters_cuda, m.def("deform_conv_backward_parameters_cuda", &deform_conv_backward_parameters_cuda,
"deform_conv_backward_parameters (CUDA)"); "deform_conv_backward_parameters (CUDA)");
m.def("modulated_deform_conv_cuda_forward", &modulated_deform_conv_cuda_forward,
"modulated deform conv forward (CUDA)");
m.def("modulated_deform_conv_cuda_backward", &modulated_deform_conv_cuda_backward,
"modulated deform conv backward (CUDA)");
} }
mmdet/ops/dcn/src/modulated_dcn_cuda.cpp → mmdet/ops/dcn/src/deform_pool_cuda.cpp
View file @ 03c38bdc
...@@ -8,32 +8,6 @@ ...@@ -8,32 +8,6 @@
#include <cmath> #include <cmath>
#include <vector> #include <vector>
void modulated_deformable_im2col_cuda(const at::Tensor data_im, const at::Tensor data_offset,
const at::Tensor data_mask, const int batch_size, const int channels,
const int height_im, const int width_im, const int height_col,
const int width_col, const int kernel_h, const int kenerl_w,
const int pad_h, const int pad_w, const int stride_h, const int stride_w,
const int dilation_h, const int dilation_w,
const int deformable_group, at::Tensor data_col);
void modulated_deformable_col2im_cuda(const at::Tensor data_col, const at::Tensor data_offset,
const at::Tensor data_mask, const int batch_size, const int channels,
const int height_im, const int width_im, const int height_col,
const int width_col, const int kernel_h, const int kenerl_w,
const int pad_h, const int pad_w, const int stride_h, const int stride_w,
const int dilation_h, const int dilation_w,
const int deformable_group, at::Tensor grad_im);
void modulated_deformable_col2im_coord_cuda(const at::Tensor data_col, const at::Tensor data_im,
const at::Tensor data_offset, const at::Tensor data_mask,
const int batch_size, const int channels, const int height_im,
const int width_im, const int height_col, const int width_col,
const int kernel_h, const int kenerl_w, const int pad_h,
const int pad_w, const int stride_h, const int stride_w,
const int dilation_h, const int dilation_w,
const int deformable_group, at::Tensor grad_offset,
at::Tensor grad_mask);
void DeformablePSROIPoolForward(const at::Tensor data, void DeformablePSROIPoolForward(const at::Tensor data,
const at::Tensor bbox, const at::Tensor bbox,
const at::Tensor trans, const at::Tensor trans,
...@@ -76,148 +50,6 @@ void DeformablePSROIPoolBackwardAcc(const at::Tensor out_grad, ...@@ -76,148 +50,6 @@ void DeformablePSROIPoolBackwardAcc(const at::Tensor out_grad,
const int sample_per_part, const int sample_per_part,
const float trans_std); const float trans_std);
void modulated_deform_conv_cuda_forward(at::Tensor input, at::Tensor weight,
at::Tensor bias, at::Tensor ones,
at::Tensor offset, at::Tensor mask,
at::Tensor output, at::Tensor columns,
int kernel_h, int kernel_w,
const int stride_h, const int stride_w,
const int pad_h, const int pad_w,
const int dilation_h, const int dilation_w,
const int deformable_group)
{
AT_CHECK(input.is_contiguous(), "input tensor has to be contiguous");
AT_CHECK(weight.is_contiguous(), "weight tensor has to be contiguous");
const int batch = input.size(0);
const int channels = input.size(1);
const int height = input.size(2);
const int width = input.size(3);
const int channels_out = weight.size(0);
const int channels_kernel = weight.size(1);
const int kernel_h_ = weight.size(2);
const int kernel_w_ = weight.size(3);
if (kernel_h_ != kernel_h || kernel_w_ != kernel_w)
AT_ERROR("Input shape and kernel shape wont match: (%d x %d vs %d x %d).",
kernel_h_, kernel_w, kernel_h_, kernel_w_);
if (channels != channels_kernel)
AT_ERROR("Input shape and kernel channels wont match: (%d vs %d).",
channels, channels_kernel);
const int height_out = (height + 2 * pad_h - (dilation_h * (kernel_h - 1) + 1)) / stride_h + 1;
const int width_out = (width + 2 * pad_w - (dilation_w * (kernel_w - 1) + 1)) / stride_w + 1;
if (ones.ndimension() != 2 ||
ones.size(0) * ones.size(1) < height_out * width_out)
{
// Resize plane and fill with ones...
ones = at::ones({height_out, width_out}, input.type());
}
// resize output
output = output.view({batch, channels_out, height_out, width_out});
// resize temporary columns
columns = at::zeros({channels * kernel_h * kernel_w, 1 * height_out * width_out}, input.type());
for (int b = 0; b < batch; b++)
{
// Do Bias first:
// M,N,K are dims of matrix A and B
// (see http://docs.nvidia.com/cuda/cublas/#cublas-lt-t-gt-gemm)
// (N x 1) (1 x M)
output[b] = output[b].flatten(1).addmm_(bias.view({-1, 1}), ones.view({1, -1}), 0.0f, 1.0f).view_as(output[b]);
modulated_deformable_im2col_cuda(input[b], offset[b], mask[b],
1, channels, height, width,
height_out, width_out, kernel_h, kernel_w,
pad_h, pad_w, stride_h, stride_w, dilation_h, dilation_w,
deformable_group, columns);
//(k * m) x (m * n)
// Y = WC
output[b] = output[b].flatten(1).addmm_(weight.flatten(1), columns).view_as(output[b]);
}
}
void modulated_deform_conv_cuda_backward(at::Tensor input, at::Tensor weight,
at::Tensor bias, at::Tensor ones,
at::Tensor offset, at::Tensor mask,
at::Tensor columns,
at::Tensor grad_input, at::Tensor grad_weight,
at::Tensor grad_bias, at::Tensor grad_offset,
at::Tensor grad_mask, at::Tensor grad_output,
int kernel_h, int kernel_w,
int stride_h, int stride_w,
int pad_h, int pad_w,
int dilation_h, int dilation_w,
int deformable_group)
{
AT_CHECK(input.is_contiguous(), "input tensor has to be contiguous");
AT_CHECK(weight.is_contiguous(), "weight tensor has to be contiguous");
const int batch = input.size(0);
const int channels = input.size(1);
const int height = input.size(2);
const int width = input.size(3);
const int channels_kernel = weight.size(1);
const int kernel_h_ = weight.size(2);
const int kernel_w_ = weight.size(3);
if (kernel_h_ != kernel_h || kernel_w_ != kernel_w)
AT_ERROR("Input shape and kernel shape wont match: (%d x %d vs %d x %d).",
kernel_h_, kernel_w, kernel_h_, kernel_w_);
if (channels != channels_kernel)
AT_ERROR("Input shape and kernel channels wont match: (%d vs %d).",
channels, channels_kernel);
const int height_out = (height + 2 * pad_h - (dilation_h * (kernel_h - 1) + 1)) / stride_h + 1;
const int width_out = (width + 2 * pad_w - (dilation_w * (kernel_w - 1) + 1)) / stride_w + 1;
if (ones.ndimension() != 2 ||
ones.size(0) * ones.size(1) < height_out * width_out)
{
// Resize plane and fill with ones...
ones = at::ones({height_out, width_out}, input.type());
}
grad_input = grad_input.view({batch, channels, height, width});
columns = at::zeros({channels * kernel_h * kernel_w, height_out * width_out}, input.type());
for (int b = 0; b < batch; b++)
{
columns.addmm_(weight.flatten(1).transpose(0, 1), grad_output[b].flatten(1), 0.0f, 1.0f);
// gradient w.r.t. input coordinate data
modulated_deformable_col2im_coord_cuda(columns, input[b], offset[b], mask[b],
1, channels, height, width,
height_out, width_out, kernel_h, kernel_w,
pad_h, pad_w, stride_h, stride_w,
dilation_h, dilation_w, deformable_group,
grad_offset[b], grad_mask[b]);
// gradient w.r.t. input data
modulated_deformable_col2im_cuda(columns, offset[b], mask[b],
1, channels, height, width,
height_out, width_out, kernel_h, kernel_w,
pad_h, pad_w, stride_h, stride_w,
dilation_h, dilation_w, deformable_group,
grad_input[b]);
// gradient w.r.t. weight, dWeight should accumulate across the batch and group
modulated_deformable_im2col_cuda(input[b], offset[b], mask[b],
1, channels, height, width,
height_out, width_out, kernel_h, kernel_w,
pad_h, pad_w, stride_h, stride_w,
dilation_h, dilation_w, deformable_group,
columns);
grad_weight = grad_weight.flatten(1).addmm_(grad_output[b].flatten(1), columns.transpose(0, 1)).view_as(grad_weight);
grad_bias = grad_bias.view({-1, 1}).addmm_(grad_output[b].flatten(1), ones.view({-1, 1})).view(-1);
}
}
void deform_psroi_pooling_cuda_forward(at::Tensor input, at::Tensor bbox, void deform_psroi_pooling_cuda_forward(at::Tensor input, at::Tensor bbox,
at::Tensor trans, at::Tensor trans,
at::Tensor out, at::Tensor top_count, at::Tensor out, at::Tensor top_count,
...@@ -305,10 +137,6 @@ void deform_psroi_pooling_cuda_backward(at::Tensor out_grad, ...@@ -305,10 +137,6 @@ void deform_psroi_pooling_cuda_backward(at::Tensor out_grad,
PYBIND11_MODULE(TORCH_EXTENSION_NAME, m) PYBIND11_MODULE(TORCH_EXTENSION_NAME, m)
{ {
m.def("modulated_deform_conv_cuda_forward", &modulated_deform_conv_cuda_forward,
"modulated deform conv forward (CUDA)");
m.def("modulated_deform_conv_cuda_backward", &modulated_deform_conv_cuda_backward,
"modulated deform conv backward (CUDA)");
m.def("deform_psroi_pooling_cuda_forward", &deform_psroi_pooling_cuda_forward, m.def("deform_psroi_pooling_cuda_forward", &deform_psroi_pooling_cuda_forward,
"deform psroi pooling forward(CUDA)"); "deform psroi pooling forward(CUDA)");
m.def("deform_psroi_pooling_cuda_backward", &deform_psroi_pooling_cuda_backward, m.def("deform_psroi_pooling_cuda_backward", &deform_psroi_pooling_cuda_backward,
......
mmdet/ops/dcn/src/deform_psroi_pooling_cuda.cu → mmdet/ops/dcn/src/deform_pool_cuda_kernel.cu
View file @ 03c38bdc
...@@ -289,18 +289,18 @@ void DeformablePSROIPoolForward(const at::Tensor data, ...@@ -289,18 +289,18 @@ void DeformablePSROIPoolForward(const at::Tensor data,
const int channels_each_class = no_trans ? output_dim : output_dim / num_classes; const int channels_each_class = no_trans ? output_dim : output_dim / num_classes;
AT_DISPATCH_FLOATING_TYPES_AND_HALF( AT_DISPATCH_FLOATING_TYPES_AND_HALF(
data.type(), "deformable_psroi_pool_forward", ([&] { data.type(), "deformable_psroi_pool_forward", ([&] {
const scalar_t *bottom_data = data.data<scalar_t>(); const scalar_t *bottom_data = data.data<scalar_t>();
const scalar_t *bottom_rois = bbox.data<scalar_t>(); const scalar_t *bottom_rois = bbox.data<scalar_t>();
const scalar_t *bottom_trans = no_trans ? NULL : trans.data<scalar_t>(); const scalar_t *bottom_trans = no_trans ? NULL : trans.data<scalar_t>();
scalar_t *top_data = out.data<scalar_t>(); scalar_t *top_data = out.data<scalar_t>();
scalar_t *top_count_data = top_count.data<scalar_t>(); scalar_t *top_count_data = top_count.data<scalar_t>();
DeformablePSROIPoolForwardKernel<<<GET_BLOCKS(count), CUDA_NUM_THREADS>>>( DeformablePSROIPoolForwardKernel<<<GET_BLOCKS(count), CUDA_NUM_THREADS>>>(
count, bottom_data, (scalar_t)spatial_scale, channels, height, width, pooled_height, pooled_width, count, bottom_data, (scalar_t)spatial_scale, channels, height, width, pooled_height, pooled_width,
bottom_rois, bottom_trans, no_trans, (scalar_t)trans_std, sample_per_part, output_dim, bottom_rois, bottom_trans, no_trans, (scalar_t)trans_std, sample_per_part, output_dim,
group_size, part_size, num_classes, channels_each_class, top_data, top_count_data); group_size, part_size, num_classes, channels_each_class, top_data, top_count_data);
})); }));
cudaError_t err = cudaGetLastError(); cudaError_t err = cudaGetLastError();
if (err != cudaSuccess) if (err != cudaSuccess)
...@@ -356,7 +356,6 @@ void DeformablePSROIPoolBackwardAcc(const at::Tensor out_grad, ...@@ -356,7 +356,6 @@ void DeformablePSROIPoolBackwardAcc(const at::Tensor out_grad,
group_size, part_size, num_classes, channels_each_class); group_size, part_size, num_classes, channels_each_class);
})); }));
cudaError_t err = cudaGetLastError(); cudaError_t err = cudaGetLastError();
if (err != cudaSuccess) if (err != cudaSuccess)
{ {
......
mmdet/ops/dcn/src/modulated_deform_im2col_cuda.cu deleted 100644 → 0
View file @ 66489d6b
This diff is collapsed.
Markdown is supported
0% or .
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment