Python torch.addcmul() Examples
The following are 30
code examples of torch.addcmul().
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
torch
, or try the search function
.
.
Example #1
| Source File: test_forward.py From incubator-tvm with Apache License 2.0 | 6 votes |
|
def test_forward_addcmul():
torch.set_grad_enabled(False)
class Addcmul1(Module):
def forward(self, *args):
t1 = torch.ones([3, 1])
t2 = torch.ones([1, 3])
if torch.cuda.is_available():
t1 = t1.cuda()
t2 = t2.cuda()
return torch.addcmul(args[0], 0.1, t1, t2)
class Addcmul2(Module):
def forward(self, *args):
return torch.addcmul(args[0], 0.5, args[1], args[2])
input_data = torch.rand([1, 3]).float()
verify_model(Addcmul1().float().eval(), input_data=input_data)
t1 = torch.rand([3, 1]).float()
t2 = torch.rand([1, 3]).float()
verify_model(Addcmul2().float().eval(), input_data=[input_data, t1, t2])
Example #2
| Source File: ElementWisedMultiply3D.py From NeuronBlocks with MIT License | 6 votes |
|
def forward(self, *args):
""" process input
Args:
*args: (Tensor): string, string_len, string2, string2_len
e.g. string (Tensor): [batch_size, seq_len, dim], string_len (Tensor): [batch_size]
Returns:
Tensor: [batch_size, seq_len, output_dim], [batch_size]
"""
dim_flag = True
input_dims = list(self.layer_conf.input_dims)
if (args[0].shape[1] * args[0].shape[2]) != (args[2].shape[1] * args[2].shape[2]):
if args[0].shape[1] == args[2].shape[1] and (input_dims[1][-1] == 1 or input_dims[0][-1] == 1):
dim_flag = True
else:
dim_flag = False
if dim_flag == False:
raise ConfigurationError("For layer ElementWisedMultiply3D, the dimensions of each inputs should be equal or 1 ,or the elements number of two inputs (expect for the first dimension) should be equal")
return torch.addcmul(torch.zeros(args[0].size()).to('cuda'),1,args[0],args[2]),args[1]
Example #3
| Source File: linear_cg.py From gpytorch with MIT License | 6 votes |
|
def _jit_linear_cg_updates(
result, alpha, residual_inner_prod, eps, beta, residual, precond_residual, mul_storage, is_zero, curr_conjugate_vec
):
# # Update result
# # result_{k} = result_{k-1} + alpha_{k} p_vec_{k-1}
result = torch.addcmul(result, alpha, curr_conjugate_vec, out=result)
# beta_{k} = (precon_residual{k}^T r_vec_{k}) / (precon_residual{k-1}^T r_vec_{k-1})
beta.resize_as_(residual_inner_prod).copy_(residual_inner_prod)
torch.mul(residual, precond_residual, out=mul_storage)
torch.sum(mul_storage, -2, keepdim=True, out=residual_inner_prod)
# Do a safe division here
torch.lt(beta, eps, out=is_zero)
beta.masked_fill_(is_zero, 1)
torch.div(residual_inner_prod, beta, out=beta)
beta.masked_fill_(is_zero, 0)
# Update curr_conjugate_vec
# curr_conjugate_vec_{k} = precon_residual{k} + beta_{k} curr_conjugate_vec_{k-1}
curr_conjugate_vec.mul_(beta).add_(precond_residual)
Example #4
| Source File: BASS.py From BASS with MIT License | 6 votes |
|
def E_Step(X, logdet, c1_temp, pi_temp, SigmaXY, X_C_SIGMA, sum, c_idx, c_idx_9, c_idx_25, distances2, r_ik_5, neig, sumP, X_C, X_C_SIGMA_buf):
"""
Computes the distances of the Data points for each centroid and normalize it,
"""
torch.add(X.unsqueeze(1), torch.neg(c1_temp.reshape(-1, Global.neig_num, Global.D_)),out=X_C)
torch.mul(X_C[:, :, 0].unsqueeze(2), SigmaXY[:, :, 0:2],out=X_C_SIGMA_buf)
torch.addcmul(X_C_SIGMA_buf,1,X_C[:,:,1].unsqueeze(2),SigmaXY[:,:,2:4],out=X_C_SIGMA[:,:,0:2])
X_C_SIGMA[:, :, 2:] = torch.mul(X_C[:, :, 2:], Global.SIGMA_INT)
torch.mul(-X_C.view(-1, Global.neig_num,Global.D_),X_C_SIGMA.view(-1,Global.neig_num,Global.D_),out=distances2)
distances2=distances2.view(-1,Global.neig_num,Global.D_)
torch.sum(distances2,2,out=r_ik_5)
r_ik_5.add_(torch.neg(logdet.reshape(-1, Global.neig_num)))
r_ik_5.add_(torch.log(pi_temp.reshape(-1, Global.neig_num)))
c_neig = c_idx_25.reshape(-1, Global.potts_area).float()
torch.add(c_neig.unsqueeze(1), -c_idx.reshape(-1, Global.neig_num).unsqueeze(2).float(),out=neig)
torch.sum((neig!=0).float(),2,out=sumP)
r_ik_5.add_(-(Global.Beta_P*sumP))
(my_help.softmaxTF(r_ik_5, 1,sum))
Example #5
| Source File: curvature.py From pytorch-sso with MIT License | 5 votes |
|
def sample_params(self, params, mean, std_scale):
for p, m, std in zip(params, mean, self.std):
noise = torch.randn_like(m)
p.data.copy_(torch.addcmul(m, std_scale, noise, std))
Example #6
| Source File: transforms.py From CenterNet with Apache License 2.0 | 5 votes |
|
def delta2bbox(rois,
deltas,
means=[0, 0, 0, 0],
stds=[1, 1, 1, 1],
max_shape=None,
wh_ratio_clip=16 / 1000):
means = deltas.new_tensor(means).repeat(1, deltas.size(1) // 4)
stds = deltas.new_tensor(stds).repeat(1, deltas.size(1) // 4)
denorm_deltas = deltas * stds + means
dx = denorm_deltas[:, 0::4]
dy = denorm_deltas[:, 1::4]
dw = denorm_deltas[:, 2::4]
dh = denorm_deltas[:, 3::4]
max_ratio = np.abs(np.log(wh_ratio_clip))
dw = dw.clamp(min=-max_ratio, max=max_ratio)
dh = dh.clamp(min=-max_ratio, max=max_ratio)
px = ((rois[:, 0] + rois[:, 2]) * 0.5).unsqueeze(1).expand_as(dx)
py = ((rois[:, 1] + rois[:, 3]) * 0.5).unsqueeze(1).expand_as(dy)
pw = (rois[:, 2] - rois[:, 0] + 1.0).unsqueeze(1).expand_as(dw)
ph = (rois[:, 3] - rois[:, 1] + 1.0).unsqueeze(1).expand_as(dh)
gw = pw * dw.exp()
gh = ph * dh.exp()
gx = torch.addcmul(px, 1, pw, dx) # gx = px + pw * dx
gy = torch.addcmul(py, 1, ph, dy) # gy = py + ph * dy
x1 = gx - gw * 0.5 + 0.5
y1 = gy - gh * 0.5 + 0.5
x2 = gx + gw * 0.5 - 0.5
y2 = gy + gh * 0.5 - 0.5
if max_shape is not None:
x1 = x1.clamp(min=0, max=max_shape[1] - 1)
y1 = y1.clamp(min=0, max=max_shape[0] - 1)
x2 = x2.clamp(min=0, max=max_shape[1] - 1)
y2 = y2.clamp(min=0, max=max_shape[0] - 1)
bboxes = torch.stack([x1, y1, x2, y2], dim=-1).view_as(deltas)
return bboxes
Example #7
| Source File: __init__.py From pytorch_compact_bilinear_pooling with BSD 3-Clause "New" or "Revised" License | 5 votes |
|
def ComplexMultiply_backward(X_re, X_im, Y_re, Y_im, grad_Z_re, grad_Z_im):
grad_X_re = torch.addcmul(grad_Z_re * Y_re, 1, grad_Z_im, Y_im)
grad_X_im = torch.addcmul(grad_Z_im * Y_re, -1, grad_Z_re, Y_im)
grad_Y_re = torch.addcmul(grad_Z_re * X_re, 1, grad_Z_im, X_im)
grad_Y_im = torch.addcmul(grad_Z_im * X_re, -1, grad_Z_re, X_im)
return grad_X_re,grad_X_im,grad_Y_re,grad_Y_im
Example #8
| Source File: __init__.py From pytorch_compact_bilinear_pooling with BSD 3-Clause "New" or "Revised" License | 5 votes |
|
def ComplexMultiply_forward(X_re, X_im, Y_re, Y_im):
Z_re = torch.addcmul(X_re*Y_re, -1, X_im, Y_im)
Z_im = torch.addcmul(X_re*Y_im, 1, X_im, Y_re)
return Z_re,Z_im
Example #9
| Source File: transforms.py From hrnet with MIT License | 5 votes |
|
def delta2bbox(rois,
deltas,
means=[0, 0, 0, 0],
stds=[1, 1, 1, 1],
max_shape=None,
wh_ratio_clip=16 / 1000):
means = deltas.new_tensor(means).repeat(1, deltas.size(1) // 4)
stds = deltas.new_tensor(stds).repeat(1, deltas.size(1) // 4)
denorm_deltas = deltas * stds + means
dx = denorm_deltas[:, 0::4]
dy = denorm_deltas[:, 1::4]
dw = denorm_deltas[:, 2::4]
dh = denorm_deltas[:, 3::4]
max_ratio = np.abs(np.log(wh_ratio_clip))
dw = dw.clamp(min=-max_ratio, max=max_ratio)
dh = dh.clamp(min=-max_ratio, max=max_ratio)
px = ((rois[:, 0] + rois[:, 2]) * 0.5).unsqueeze(1).expand_as(dx)
py = ((rois[:, 1] + rois[:, 3]) * 0.5).unsqueeze(1).expand_as(dy)
pw = (rois[:, 2] - rois[:, 0] + 1.0).unsqueeze(1).expand_as(dw)
ph = (rois[:, 3] - rois[:, 1] + 1.0).unsqueeze(1).expand_as(dh)
gw = pw * dw.exp()
gh = ph * dh.exp()
gx = torch.addcmul(px, 1, pw, dx) # gx = px + pw * dx
gy = torch.addcmul(py, 1, ph, dy) # gy = py + ph * dy
x1 = gx - gw * 0.5 + 0.5
y1 = gy - gh * 0.5 + 0.5
x2 = gx + gw * 0.5 - 0.5
y2 = gy + gh * 0.5 - 0.5
if max_shape is not None:
x1 = x1.clamp(min=0, max=max_shape[1] - 1)
y1 = y1.clamp(min=0, max=max_shape[0] - 1)
x2 = x2.clamp(min=0, max=max_shape[1] - 1)
y2 = y2.clamp(min=0, max=max_shape[0] - 1)
bboxes = torch.stack([x1, y1, x2, y2], dim=-1).view_as(deltas)
return bboxes
Example #10
| Source File: gradcam.py From pytorch-grad-cam with MIT License | 5 votes |
|
def backward(self, grad_output):
input, output = self.saved_tensors
grad_input = None
positive_mask_1 = (input > 0).type_as(grad_output)
positive_mask_2 = (grad_output > 0).type_as(grad_output)
grad_input = torch.addcmul(torch.zeros(input.size()).type_as(input),
torch.addcmul(torch.zeros(input.size()).type_as(input), grad_output,
positive_mask_1), positive_mask_2)
return grad_input
Example #11
| Source File: gradcam.py From pytorch-grad-cam with MIT License | 5 votes |
|
def forward(self, input):
positive_mask = (input > 0).type_as(input)
output = torch.addcmul(torch.zeros(input.size()).type_as(input), input, positive_mask)
self.save_for_backward(input, output)
return output
Example #12
| Source File: transforms.py From mmaction with Apache License 2.0 | 5 votes |
|
def delta2bbox(rois,
deltas,
means=[0, 0, 0, 0],
stds=[1, 1, 1, 1],
max_shape=None,
wh_ratio_clip=16 / 1000):
means = deltas.new_tensor(means).repeat(1, deltas.size(1) // 4)
stds = deltas.new_tensor(stds).repeat(1, deltas.size(1) // 4)
denorm_deltas = deltas * stds + means
dx = denorm_deltas[:, 0::4]
dy = denorm_deltas[:, 1::4]
dw = denorm_deltas[:, 2::4]
dh = denorm_deltas[:, 3::4]
max_ratio = np.abs(np.log(wh_ratio_clip))
dw = dw.clamp(min=-max_ratio, max=max_ratio)
dh = dh.clamp(min=-max_ratio, max=max_ratio)
px = ((rois[:, 0] + rois[:, 2]) * 0.5).unsqueeze(1).expand_as(dx)
py = ((rois[:, 1] + rois[:, 3]) * 0.5).unsqueeze(1).expand_as(dy)
pw = ((rois[:, 2] - rois[:, 0]) + 1.0).unsqueeze(1).expand_as(dw)
ph = ((rois[:, 3] - rois[:, 1]) + 1.0).unsqueeze(1).expand_as(dh)
gw = pw * dw.exp()
gh = ph * dh.exp()
gx = torch.addcmul(px, 1, pw, dx) # gx = px + pw * dx
gy = torch.addcmul(py, 1, ph, dy) # gy = py + ph * dy
x1 = gx - gw * 0.5 + 0.5
y1 = gy - gh * 0.5 + 0.5
x2 = gx + gw * 0.5 - 0.5
y2 = gy + gh * 0.5 - 0.5
if max_shape is not None:
x1 = x1.clamp(min=0, max=max_shape[1] - 1)
y1 = y1.clamp(min=0, max=max_shape[0] - 1)
x2 = x2.clamp(min=0, max=max_shape[1] - 1)
y2 = y2.clamp(min=0, max=max_shape[0] - 1)
bboxes = torch.stack([x1, y1, x2, y2], dim=-1).view_as(deltas)
return bboxes
Example #13
| Source File: transforms.py From kaggle-imaterialist with MIT License | 5 votes |
|
def delta2bbox(rois,
deltas,
means=[0, 0, 0, 0],
stds=[1, 1, 1, 1],
max_shape=None,
wh_ratio_clip=16 / 1000):
means = deltas.new_tensor(means).repeat(1, deltas.size(1) // 4)
stds = deltas.new_tensor(stds).repeat(1, deltas.size(1) // 4)
denorm_deltas = deltas * stds + means
dx = denorm_deltas[:, 0::4]
dy = denorm_deltas[:, 1::4]
dw = denorm_deltas[:, 2::4]
dh = denorm_deltas[:, 3::4]
max_ratio = np.abs(np.log(wh_ratio_clip))
dw = dw.clamp(min=-max_ratio, max=max_ratio)
dh = dh.clamp(min=-max_ratio, max=max_ratio)
px = ((rois[:, 0] + rois[:, 2]) * 0.5).unsqueeze(1).expand_as(dx)
py = ((rois[:, 1] + rois[:, 3]) * 0.5).unsqueeze(1).expand_as(dy)
pw = (rois[:, 2] - rois[:, 0] + 1.0).unsqueeze(1).expand_as(dw)
ph = (rois[:, 3] - rois[:, 1] + 1.0).unsqueeze(1).expand_as(dh)
gw = pw * dw.exp()
gh = ph * dh.exp()
gx = torch.addcmul(px, 1, pw, dx) # gx = px + pw * dx
gy = torch.addcmul(py, 1, ph, dy) # gy = py + ph * dy
x1 = gx - gw * 0.5 + 0.5
y1 = gy - gh * 0.5 + 0.5
x2 = gx + gw * 0.5 - 0.5
y2 = gy + gh * 0.5 - 0.5
if max_shape is not None:
x1 = x1.clamp(min=0, max=max_shape[1] - 1)
y1 = y1.clamp(min=0, max=max_shape[0] - 1)
x2 = x2.clamp(min=0, max=max_shape[1] - 1)
y2 = y2.clamp(min=0, max=max_shape[0] - 1)
bboxes = torch.stack([x1, y1, x2, y2], dim=-1).view_as(deltas)
return bboxes
Example #14
| Source File: transforms.py From Feature-Selective-Anchor-Free-Module-for-Single-Shot-Object-Detection with Apache License 2.0 | 5 votes |
|
def delta2bbox(rois,
deltas,
means=[0, 0, 0, 0],
stds=[1, 1, 1, 1],
max_shape=None,
wh_ratio_clip=16 / 1000):
means = deltas.new_tensor(means).repeat(1, deltas.size(1) // 4)
stds = deltas.new_tensor(stds).repeat(1, deltas.size(1) // 4)
denorm_deltas = deltas * stds + means
dx = denorm_deltas[:, 0::4]
dy = denorm_deltas[:, 1::4]
dw = denorm_deltas[:, 2::4]
dh = denorm_deltas[:, 3::4]
max_ratio = np.abs(np.log(wh_ratio_clip))
dw = dw.clamp(min=-max_ratio, max=max_ratio)
dh = dh.clamp(min=-max_ratio, max=max_ratio)
px = ((rois[:, 0] + rois[:, 2]) * 0.5).unsqueeze(1).expand_as(dx)
py = ((rois[:, 1] + rois[:, 3]) * 0.5).unsqueeze(1).expand_as(dy)
pw = (rois[:, 2] - rois[:, 0] + 1.0).unsqueeze(1).expand_as(dw)
ph = (rois[:, 3] - rois[:, 1] + 1.0).unsqueeze(1).expand_as(dh)
gw = pw * dw.exp()
gh = ph * dh.exp()
gx = torch.addcmul(px, 1, pw, dx) # gx = px + pw * dx
gy = torch.addcmul(py, 1, ph, dy) # gy = py + ph * dy
x1 = gx - gw * 0.5 + 0.5
y1 = gy - gh * 0.5 + 0.5
x2 = gx + gw * 0.5 - 0.5
y2 = gy + gh * 0.5 - 0.5
if max_shape is not None:
x1 = x1.clamp(min=0, max=max_shape[1] - 1)
y1 = y1.clamp(min=0, max=max_shape[0] - 1)
x2 = x2.clamp(min=0, max=max_shape[1] - 1)
y2 = y2.clamp(min=0, max=max_shape[0] - 1)
bboxes = torch.stack([x1, y1, x2, y2], dim=-1).view_as(deltas)
return bboxes
Example #15
| Source File: transforms.py From Cascade-RPN with Apache License 2.0 | 5 votes |
|
def delta2bbox(rois,
deltas,
means=[0, 0, 0, 0],
stds=[1, 1, 1, 1],
max_shape=None,
wh_ratio_clip=16 / 1000):
means = deltas.new_tensor(means).repeat(1, deltas.size(1) // 4)
stds = deltas.new_tensor(stds).repeat(1, deltas.size(1) // 4)
denorm_deltas = deltas * stds + means
dx = denorm_deltas[:, 0::4]
dy = denorm_deltas[:, 1::4]
dw = denorm_deltas[:, 2::4]
dh = denorm_deltas[:, 3::4]
max_ratio = np.abs(np.log(wh_ratio_clip))
dw = dw.clamp(min=-max_ratio, max=max_ratio)
dh = dh.clamp(min=-max_ratio, max=max_ratio)
px = ((rois[:, 0] + rois[:, 2]) * 0.5).unsqueeze(1).expand_as(dx)
py = ((rois[:, 1] + rois[:, 3]) * 0.5).unsqueeze(1).expand_as(dy)
pw = (rois[:, 2] - rois[:, 0] + 1.0).unsqueeze(1).expand_as(dw)
ph = (rois[:, 3] - rois[:, 1] + 1.0).unsqueeze(1).expand_as(dh)
gw = pw * dw.exp()
gh = ph * dh.exp()
gx = torch.addcmul(px, 1, pw, dx) # gx = px + pw * dx
gy = torch.addcmul(py, 1, ph, dy) # gy = py + ph * dy
x1 = gx - gw * 0.5 + 0.5
y1 = gy - gh * 0.5 + 0.5
x2 = gx + gw * 0.5 - 0.5
y2 = gy + gh * 0.5 - 0.5
if max_shape is not None:
x1 = x1.clamp(min=0, max=max_shape[1] - 1)
y1 = y1.clamp(min=0, max=max_shape[0] - 1)
x2 = x2.clamp(min=0, max=max_shape[1] - 1)
y2 = y2.clamp(min=0, max=max_shape[0] - 1)
bboxes = torch.stack([x1, y1, x2, y2], dim=-1).view_as(deltas)
return bboxes
Example #16
| Source File: transforms.py From AugFPN with Apache License 2.0 | 5 votes |
|
def delta2bbox(rois,
deltas,
means=[0, 0, 0, 0],
stds=[1, 1, 1, 1],
max_shape=None,
wh_ratio_clip=16 / 1000):
means = deltas.new_tensor(means).repeat(1, deltas.size(1) // 4)
stds = deltas.new_tensor(stds).repeat(1, deltas.size(1) // 4)
denorm_deltas = deltas * stds + means
dx = denorm_deltas[:, 0::4]
dy = denorm_deltas[:, 1::4]
dw = denorm_deltas[:, 2::4]
dh = denorm_deltas[:, 3::4]
max_ratio = np.abs(np.log(wh_ratio_clip))
dw = dw.clamp(min=-max_ratio, max=max_ratio)
dh = dh.clamp(min=-max_ratio, max=max_ratio)
px = ((rois[:, 0] + rois[:, 2]) * 0.5).unsqueeze(1).expand_as(dx)
py = ((rois[:, 1] + rois[:, 3]) * 0.5).unsqueeze(1).expand_as(dy)
pw = (rois[:, 2] - rois[:, 0] + 1.0).unsqueeze(1).expand_as(dw)
ph = (rois[:, 3] - rois[:, 1] + 1.0).unsqueeze(1).expand_as(dh)
gw = pw * dw.exp()
gh = ph * dh.exp()
gx = torch.addcmul(px, 1, pw, dx) # gx = px + pw * dx
gy = torch.addcmul(py, 1, ph, dy) # gy = py + ph * dy
x1 = gx - gw * 0.5 + 0.5
y1 = gy - gh * 0.5 + 0.5
x2 = gx + gw * 0.5 - 0.5
y2 = gy + gh * 0.5 - 0.5
if max_shape is not None:
x1 = x1.clamp(min=0, max=max_shape[1] - 1)
y1 = y1.clamp(min=0, max=max_shape[0] - 1)
x2 = x2.clamp(min=0, max=max_shape[1] - 1)
y2 = y2.clamp(min=0, max=max_shape[0] - 1)
bboxes = torch.stack([x1, y1, x2, y2], dim=-1).view_as(deltas)
return bboxes
Example #17
| Source File: transforms.py From FoveaBox with Apache License 2.0 | 5 votes |
|
def delta2bbox(rois,
deltas,
means=[0, 0, 0, 0],
stds=[1, 1, 1, 1],
max_shape=None,
wh_ratio_clip=16 / 1000):
means = deltas.new_tensor(means).repeat(1, deltas.size(1) // 4)
stds = deltas.new_tensor(stds).repeat(1, deltas.size(1) // 4)
denorm_deltas = deltas * stds + means
dx = denorm_deltas[:, 0::4]
dy = denorm_deltas[:, 1::4]
dw = denorm_deltas[:, 2::4]
dh = denorm_deltas[:, 3::4]
max_ratio = np.abs(np.log(wh_ratio_clip))
dw = dw.clamp(min=-max_ratio, max=max_ratio)
dh = dh.clamp(min=-max_ratio, max=max_ratio)
px = ((rois[:, 0] + rois[:, 2]) * 0.5).unsqueeze(1).expand_as(dx)
py = ((rois[:, 1] + rois[:, 3]) * 0.5).unsqueeze(1).expand_as(dy)
pw = (rois[:, 2] - rois[:, 0] + 1.0).unsqueeze(1).expand_as(dw)
ph = (rois[:, 3] - rois[:, 1] + 1.0).unsqueeze(1).expand_as(dh)
gw = pw * dw.exp()
gh = ph * dh.exp()
gx = torch.addcmul(px, 1, pw, dx) # gx = px + pw * dx
gy = torch.addcmul(py, 1, ph, dy) # gy = py + ph * dy
x1 = gx - gw * 0.5 + 0.5
y1 = gy - gh * 0.5 + 0.5
x2 = gx + gw * 0.5 - 0.5
y2 = gy + gh * 0.5 - 0.5
if max_shape is not None:
x1 = x1.clamp(min=0, max=max_shape[1] - 1)
y1 = y1.clamp(min=0, max=max_shape[0] - 1)
x2 = x2.clamp(min=0, max=max_shape[1] - 1)
y2 = y2.clamp(min=0, max=max_shape[0] - 1)
bboxes = torch.stack([x1, y1, x2, y2], dim=-1).view_as(deltas)
return bboxes
Example #18
| Source File: ElementWisedMultiply2D.py From NeuronBlocks with MIT License | 5 votes |
|
def forward(self, *args):
""" process input
Args:
*args: (Tensor): string, string_len, string2, string2_len
e.g. string (Tensor): [batch_size, dim], string_len (Tensor): [batch_size]
Returns:
Tensor: [batch_size, output_dim], [batch_size]
"""
return torch.addcmul(torch.zeros(args[0].size()).to('cuda'),1,args[0],args[2]),args[1]
Example #19
| Source File: transforms.py From Libra_R-CNN with Apache License 2.0 | 5 votes |
|
def delta2bbox(rois,
deltas,
means=[0, 0, 0, 0],
stds=[1, 1, 1, 1],
max_shape=None,
wh_ratio_clip=16 / 1000):
means = deltas.new_tensor(means).repeat(1, deltas.size(1) // 4)
stds = deltas.new_tensor(stds).repeat(1, deltas.size(1) // 4)
denorm_deltas = deltas * stds + means
dx = denorm_deltas[:, 0::4]
dy = denorm_deltas[:, 1::4]
dw = denorm_deltas[:, 2::4]
dh = denorm_deltas[:, 3::4]
max_ratio = np.abs(np.log(wh_ratio_clip))
dw = dw.clamp(min=-max_ratio, max=max_ratio)
dh = dh.clamp(min=-max_ratio, max=max_ratio)
px = ((rois[:, 0] + rois[:, 2]) * 0.5).unsqueeze(1).expand_as(dx)
py = ((rois[:, 1] + rois[:, 3]) * 0.5).unsqueeze(1).expand_as(dy)
pw = (rois[:, 2] - rois[:, 0] + 1.0).unsqueeze(1).expand_as(dw)
ph = (rois[:, 3] - rois[:, 1] + 1.0).unsqueeze(1).expand_as(dh)
gw = pw * dw.exp()
gh = ph * dh.exp()
gx = torch.addcmul(px, 1, pw, dx) # gx = px + pw * dx
gy = torch.addcmul(py, 1, ph, dy) # gy = py + ph * dy
x1 = gx - gw * 0.5 + 0.5
y1 = gy - gh * 0.5 + 0.5
x2 = gx + gw * 0.5 - 0.5
y2 = gy + gh * 0.5 - 0.5
if max_shape is not None:
x1 = x1.clamp(min=0, max=max_shape[1] - 1)
y1 = y1.clamp(min=0, max=max_shape[0] - 1)
x2 = x2.clamp(min=0, max=max_shape[1] - 1)
y2 = y2.clamp(min=0, max=max_shape[0] - 1)
bboxes = torch.stack([x1, y1, x2, y2], dim=-1).view_as(deltas)
return bboxes
Example #20
| Source File: transforms.py From mmdetection_with_SENet154 with Apache License 2.0 | 5 votes |
|
def delta2bbox(rois,
deltas,
means=[0, 0, 0, 0],
stds=[1, 1, 1, 1],
max_shape=None,
wh_ratio_clip=16 / 1000):
means = deltas.new_tensor(means).repeat(1, deltas.size(1) // 4)
stds = deltas.new_tensor(stds).repeat(1, deltas.size(1) // 4)
denorm_deltas = deltas * stds + means
dx = denorm_deltas[:, 0::4]
dy = denorm_deltas[:, 1::4]
dw = denorm_deltas[:, 2::4]
dh = denorm_deltas[:, 3::4]
max_ratio = np.abs(np.log(wh_ratio_clip))
dw = dw.clamp(min=-max_ratio, max=max_ratio)
dh = dh.clamp(min=-max_ratio, max=max_ratio)
px = ((rois[:, 0] + rois[:, 2]) * 0.5).unsqueeze(1).expand_as(dx)
py = ((rois[:, 1] + rois[:, 3]) * 0.5).unsqueeze(1).expand_as(dy)
pw = (rois[:, 2] - rois[:, 0] + 1.0).unsqueeze(1).expand_as(dw)
ph = (rois[:, 3] - rois[:, 1] + 1.0).unsqueeze(1).expand_as(dh)
gw = pw * dw.exp()
gh = ph * dh.exp()
gx = torch.addcmul(px, 1, pw, dx) # gx = px + pw * dx
gy = torch.addcmul(py, 1, ph, dy) # gy = py + ph * dy
x1 = gx - gw * 0.5 + 0.5
y1 = gy - gh * 0.5 + 0.5
x2 = gx + gw * 0.5 - 0.5
y2 = gy + gh * 0.5 - 0.5
if max_shape is not None:
x1 = x1.clamp(min=0, max=max_shape[1] - 1)
y1 = y1.clamp(min=0, max=max_shape[0] - 1)
x2 = x2.clamp(min=0, max=max_shape[1] - 1)
y2 = y2.clamp(min=0, max=max_shape[0] - 1)
bboxes = torch.stack([x1, y1, x2, y2], dim=-1).view_as(deltas)
return bboxes
Example #21
| Source File: compactbilinearpooling.py From block.bootstrap.pytorch with BSD 3-Clause "New" or "Revised" License | 5 votes |
|
def ComplexMultiply_forward(X_re, X_im, Y_re, Y_im):
Z_re = torch.addcmul(X_re*Y_re, -1, X_im, Y_im)
Z_im = torch.addcmul(X_re*Y_im, 1, X_im, Y_re)
return Z_re,Z_im
Example #22
| Source File: compactbilinearpooling.py From block.bootstrap.pytorch with BSD 3-Clause "New" or "Revised" License | 5 votes |
|
def ComplexMultiply_backward(X_re, X_im, Y_re, Y_im, grad_Z_re, grad_Z_im):
grad_X_re = torch.addcmul(grad_Z_re * Y_re, 1, grad_Z_im, Y_im)
grad_X_im = torch.addcmul(grad_Z_im * Y_re, -1, grad_Z_re, Y_im)
grad_Y_re = torch.addcmul(grad_Z_re * X_re, 1, grad_Z_im, X_im)
grad_Y_im = torch.addcmul(grad_Z_im * X_re, -1, grad_Z_re, X_im)
return grad_X_re,grad_X_im,grad_Y_re,grad_Y_im
Example #23
| Source File: transforms.py From GCNet with Apache License 2.0 | 5 votes |
|
def delta2bbox(rois,
deltas,
means=[0, 0, 0, 0],
stds=[1, 1, 1, 1],
max_shape=None,
wh_ratio_clip=16 / 1000):
means = deltas.new_tensor(means).repeat(1, deltas.size(1) // 4)
stds = deltas.new_tensor(stds).repeat(1, deltas.size(1) // 4)
denorm_deltas = deltas * stds + means
dx = denorm_deltas[:, 0::4]
dy = denorm_deltas[:, 1::4]
dw = denorm_deltas[:, 2::4]
dh = denorm_deltas[:, 3::4]
max_ratio = np.abs(np.log(wh_ratio_clip))
dw = dw.clamp(min=-max_ratio, max=max_ratio)
dh = dh.clamp(min=-max_ratio, max=max_ratio)
px = ((rois[:, 0] + rois[:, 2]) * 0.5).unsqueeze(1).expand_as(dx)
py = ((rois[:, 1] + rois[:, 3]) * 0.5).unsqueeze(1).expand_as(dy)
pw = (rois[:, 2] - rois[:, 0] + 1.0).unsqueeze(1).expand_as(dw)
ph = (rois[:, 3] - rois[:, 1] + 1.0).unsqueeze(1).expand_as(dh)
gw = pw * dw.exp()
gh = ph * dh.exp()
gx = torch.addcmul(px, 1, pw, dx) # gx = px + pw * dx
gy = torch.addcmul(py, 1, ph, dy) # gy = py + ph * dy
x1 = gx - gw * 0.5 + 0.5
y1 = gy - gh * 0.5 + 0.5
x2 = gx + gw * 0.5 - 0.5
y2 = gy + gh * 0.5 - 0.5
if max_shape is not None:
x1 = x1.clamp(min=0, max=max_shape[1] - 1)
y1 = y1.clamp(min=0, max=max_shape[0] - 1)
x2 = x2.clamp(min=0, max=max_shape[1] - 1)
y2 = y2.clamp(min=0, max=max_shape[0] - 1)
bboxes = torch.stack([x1, y1, x2, y2], dim=-1).view_as(deltas)
return bboxes
Example #24
| Source File: gradients.py From pytorch-smoothgrad with MIT License | 5 votes |
|
def forward(self, input):
pos_mask = (input > 0).type_as(input)
output = torch.addcmul(
torch.zeros(input.size()).type_as(input),
input,
pos_mask)
self.save_for_backward(input, output)
return output
Example #25
| Source File: gradients.py From pytorch-smoothgrad with MIT License | 5 votes |
|
def backward(self, grad_output):
input, output = self.saved_tensors
pos_mask_1 = (input > 0).type_as(grad_output)
pos_mask_2 = (grad_output > 0).type_as(grad_output)
grad_input = torch.addcmul(
torch.zeros(input.size()).type_as(input),
torch.addcmul(
torch.zeros(input.size()).type_as(input), grad_output, pos_mask_1),
pos_mask_2)
return grad_input
Example #26
| Source File: transforms.py From mmdetection-annotated with Apache License 2.0 | 5 votes |
|
def delta2bbox(rois,
deltas,
means=[0, 0, 0, 0],
stds=[1, 1, 1, 1],
max_shape=None,
wh_ratio_clip=16 / 1000):
means = deltas.new_tensor(means).repeat(1, deltas.size(1) // 4)
stds = deltas.new_tensor(stds).repeat(1, deltas.size(1) // 4)
denorm_deltas = deltas * stds + means
dx = denorm_deltas[:, 0::4]
dy = denorm_deltas[:, 1::4]
dw = denorm_deltas[:, 2::4]
dh = denorm_deltas[:, 3::4]
max_ratio = np.abs(np.log(wh_ratio_clip))
dw = dw.clamp(min=-max_ratio, max=max_ratio)
dh = dh.clamp(min=-max_ratio, max=max_ratio)
px = ((rois[:, 0] + rois[:, 2]) * 0.5).unsqueeze(1).expand_as(dx)
py = ((rois[:, 1] + rois[:, 3]) * 0.5).unsqueeze(1).expand_as(dy)
pw = (rois[:, 2] - rois[:, 0] + 1.0).unsqueeze(1).expand_as(dw)
ph = (rois[:, 3] - rois[:, 1] + 1.0).unsqueeze(1).expand_as(dh)
gw = pw * dw.exp()
gh = ph * dh.exp()
gx = torch.addcmul(px, 1, pw, dx) # gx = px + pw * dx
gy = torch.addcmul(py, 1, ph, dy) # gy = py + ph * dy
x1 = gx - gw * 0.5 + 0.5
y1 = gy - gh * 0.5 + 0.5
x2 = gx + gw * 0.5 - 0.5
y2 = gy + gh * 0.5 - 0.5
if max_shape is not None:
x1 = x1.clamp(min=0, max=max_shape[1] - 1)
y1 = y1.clamp(min=0, max=max_shape[0] - 1)
x2 = x2.clamp(min=0, max=max_shape[1] - 1)
y2 = y2.clamp(min=0, max=max_shape[0] - 1)
bboxes = torch.stack([x1, y1, x2, y2], dim=-1).view_as(deltas)
return bboxes
Example #27
| Source File: transforms.py From AerialDetection with Apache License 2.0 | 5 votes |
|
def delta2bbox(rois,
deltas,
means=[0, 0, 0, 0],
stds=[1, 1, 1, 1],
max_shape=None,
wh_ratio_clip=16 / 1000):
means = deltas.new_tensor(means).repeat(1, deltas.size(1) // 4)
stds = deltas.new_tensor(stds).repeat(1, deltas.size(1) // 4)
denorm_deltas = deltas * stds + means
dx = denorm_deltas[:, 0::4]
dy = denorm_deltas[:, 1::4]
dw = denorm_deltas[:, 2::4]
dh = denorm_deltas[:, 3::4]
max_ratio = np.abs(np.log(wh_ratio_clip))
dw = dw.clamp(min=-max_ratio, max=max_ratio)
dh = dh.clamp(min=-max_ratio, max=max_ratio)
px = ((rois[:, 0] + rois[:, 2]) * 0.5).unsqueeze(1).expand_as(dx)
py = ((rois[:, 1] + rois[:, 3]) * 0.5).unsqueeze(1).expand_as(dy)
pw = (rois[:, 2] - rois[:, 0] + 1.0).unsqueeze(1).expand_as(dw)
ph = (rois[:, 3] - rois[:, 1] + 1.0).unsqueeze(1).expand_as(dh)
gw = pw * dw.exp()
gh = ph * dh.exp()
gx = torch.addcmul(px, 1, pw, dx) # gx = px + pw * dx
gy = torch.addcmul(py, 1, ph, dy) # gy = py + ph * dy
x1 = gx - gw * 0.5 + 0.5
y1 = gy - gh * 0.5 + 0.5
x2 = gx + gw * 0.5 - 0.5
y2 = gy + gh * 0.5 - 0.5
if max_shape is not None:
x1 = x1.clamp(min=0, max=max_shape[1] - 1)
y1 = y1.clamp(min=0, max=max_shape[0] - 1)
x2 = x2.clamp(min=0, max=max_shape[1] - 1)
y2 = y2.clamp(min=0, max=max_shape[0] - 1)
bboxes = torch.stack([x1, y1, x2, y2], dim=-1).view_as(deltas)
return bboxes
Example #28
| Source File: transforms.py From PolarMask with Apache License 2.0 | 5 votes |
|
def delta2bbox(rois,
deltas,
means=[0, 0, 0, 0],
stds=[1, 1, 1, 1],
max_shape=None,
wh_ratio_clip=16 / 1000):
means = deltas.new_tensor(means).repeat(1, deltas.size(1) // 4)
stds = deltas.new_tensor(stds).repeat(1, deltas.size(1) // 4)
denorm_deltas = deltas * stds + means
dx = denorm_deltas[:, 0::4]
dy = denorm_deltas[:, 1::4]
dw = denorm_deltas[:, 2::4]
dh = denorm_deltas[:, 3::4]
max_ratio = np.abs(np.log(wh_ratio_clip))
dw = dw.clamp(min=-max_ratio, max=max_ratio)
dh = dh.clamp(min=-max_ratio, max=max_ratio)
px = ((rois[:, 0] + rois[:, 2]) * 0.5).unsqueeze(1).expand_as(dx)
py = ((rois[:, 1] + rois[:, 3]) * 0.5).unsqueeze(1).expand_as(dy)
pw = (rois[:, 2] - rois[:, 0] + 1.0).unsqueeze(1).expand_as(dw)
ph = (rois[:, 3] - rois[:, 1] + 1.0).unsqueeze(1).expand_as(dh)
gw = pw * dw.exp()
gh = ph * dh.exp()
gx = torch.addcmul(px, 1, pw, dx) # gx = px + pw * dx
gy = torch.addcmul(py, 1, ph, dy) # gy = py + ph * dy
x1 = gx - gw * 0.5 + 0.5
y1 = gy - gh * 0.5 + 0.5
x2 = gx + gw * 0.5 - 0.5
y2 = gy + gh * 0.5 - 0.5
if max_shape is not None:
x1 = x1.clamp(min=0, max=max_shape[1] - 1)
y1 = y1.clamp(min=0, max=max_shape[0] - 1)
x2 = x2.clamp(min=0, max=max_shape[1] - 1)
y2 = y2.clamp(min=0, max=max_shape[0] - 1)
bboxes = torch.stack([x1, y1, x2, y2], dim=-1).view_as(deltas)
return bboxes
Example #29
| Source File: transforms.py From kaggle-kuzushiji-recognition with MIT License | 5 votes |
|
def delta2bbox(rois,
deltas,
means=[0, 0, 0, 0],
stds=[1, 1, 1, 1],
max_shape=None,
wh_ratio_clip=16 / 1000):
means = deltas.new_tensor(means).repeat(1, deltas.size(1) // 4)
stds = deltas.new_tensor(stds).repeat(1, deltas.size(1) // 4)
denorm_deltas = deltas * stds + means
dx = denorm_deltas[:, 0::4]
dy = denorm_deltas[:, 1::4]
dw = denorm_deltas[:, 2::4]
dh = denorm_deltas[:, 3::4]
max_ratio = np.abs(np.log(wh_ratio_clip))
dw = dw.clamp(min=-max_ratio, max=max_ratio)
dh = dh.clamp(min=-max_ratio, max=max_ratio)
px = ((rois[:, 0] + rois[:, 2]) * 0.5).unsqueeze(1).expand_as(dx)
py = ((rois[:, 1] + rois[:, 3]) * 0.5).unsqueeze(1).expand_as(dy)
pw = (rois[:, 2] - rois[:, 0] + 1.0).unsqueeze(1).expand_as(dw)
ph = (rois[:, 3] - rois[:, 1] + 1.0).unsqueeze(1).expand_as(dh)
gw = pw * dw.exp()
gh = ph * dh.exp()
gx = torch.addcmul(px, 1, pw, dx) # gx = px + pw * dx
gy = torch.addcmul(py, 1, ph, dy) # gy = py + ph * dy
x1 = gx - gw * 0.5 + 0.5
y1 = gy - gh * 0.5 + 0.5
x2 = gx + gw * 0.5 - 0.5
y2 = gy + gh * 0.5 - 0.5
if max_shape is not None:
x1 = x1.clamp(min=0, max=max_shape[1] - 1)
y1 = y1.clamp(min=0, max=max_shape[0] - 1)
x2 = x2.clamp(min=0, max=max_shape[1] - 1)
y2 = y2.clamp(min=0, max=max_shape[0] - 1)
bboxes = torch.stack([x1, y1, x2, y2], dim=-1).view_as(deltas)
return bboxes
Example #30
| Source File: transforms.py From Grid-R-CNN with Apache License 2.0 | 5 votes |
|
def delta2bbox(rois,
deltas,
means=[0, 0, 0, 0],
stds=[1, 1, 1, 1],
max_shape=None,
wh_ratio_clip=16 / 1000):
means = deltas.new_tensor(means).repeat(1, deltas.size(1) // 4)
stds = deltas.new_tensor(stds).repeat(1, deltas.size(1) // 4)
denorm_deltas = deltas * stds + means
dx = denorm_deltas[:, 0::4]
dy = denorm_deltas[:, 1::4]
dw = denorm_deltas[:, 2::4]
dh = denorm_deltas[:, 3::4]
max_ratio = np.abs(np.log(wh_ratio_clip))
dw = dw.clamp(min=-max_ratio, max=max_ratio)
dh = dh.clamp(min=-max_ratio, max=max_ratio)
px = ((rois[:, 0] + rois[:, 2]) * 0.5).unsqueeze(1).expand_as(dx)
py = ((rois[:, 1] + rois[:, 3]) * 0.5).unsqueeze(1).expand_as(dy)
pw = (rois[:, 2] - rois[:, 0] + 1.0).unsqueeze(1).expand_as(dw)
ph = (rois[:, 3] - rois[:, 1] + 1.0).unsqueeze(1).expand_as(dh)
gw = pw * dw.exp()
gh = ph * dh.exp()
gx = torch.addcmul(px, 1, pw, dx) # gx = px + pw * dx
gy = torch.addcmul(py, 1, ph, dy) # gy = py + ph * dy
x1 = gx - gw * 0.5 + 0.5
y1 = gy - gh * 0.5 + 0.5
x2 = gx + gw * 0.5 - 0.5
y2 = gy + gh * 0.5 - 0.5
if max_shape is not None:
x1 = x1.clamp(min=0, max=max_shape[1] - 1)
y1 = y1.clamp(min=0, max=max_shape[0] - 1)
x2 = x2.clamp(min=0, max=max_shape[1] - 1)
y2 = y2.clamp(min=0, max=max_shape[0] - 1)
bboxes = torch.stack([x1, y1, x2, y2], dim=-1).view_as(deltas)
return bboxes
