本文整理汇总了Python中mxnet.nd.expand_dims方法的典型用法代码示例。如果您正苦于以下问题:Python nd.expand_dims方法的具体用法?Python nd.expand_dims怎么用?Python nd.expand_dims使用的例子?那么, 这里精选的方法代码示例或许可以为您提供帮助。您也可以进一步了解该方法所在类mxnet.nd
的用法示例。
在下文中一共展示了nd.expand_dims方法的15个代码示例,这些例子默认根据受欢迎程度排序。您可以为喜欢或者感觉有用的代码点赞,您的评价将有助于系统推荐出更棒的Python代码示例。
示例1: extract_pairwise_multi_position_embedding_nd
# 需要导入模块: from mxnet import nd [as 别名]
# 或者: from mxnet.nd import expand_dims [as 别名]
def extract_pairwise_multi_position_embedding_nd(position_mat, feat_dim, wave_length=1000):
""" Extract multi-class position embedding
Args:
position_mat: [num_fg_classes, num_rois, num_rois, 4]
feat_dim: dimension of embedding feature
wave_length:
Returns:
embedding: [num_fg_classes, num_rois, num_rois, feat_dim]
"""
feat_range = nd.arange(0, feat_dim / 8)
dim_mat = nd.broadcast_power(lhs=nd.full((1,), wave_length),
rhs=(8. / feat_dim) * feat_range)
dim_mat = nd.Reshape(dim_mat, shape=(1, 1, 1, 1, -1))
position_mat = nd.expand_dims(100.0 * position_mat, axis=4)
div_mat = nd.broadcast_div(lhs=position_mat, rhs=dim_mat)
sin_mat = nd.sin(data=div_mat)
cos_mat = nd.cos(data=div_mat)
# embedding, [num_fg_classes, num_rois, num_rois, 4, feat_dim/4]
embedding = nd.concat(sin_mat, cos_mat, dim=4)
embedding = nd.Reshape(embedding, shape=(0, 0, 0, feat_dim))
return embedding
示例2: Route
# 需要导入模块: from mxnet import nd [as 别名]
# 或者: from mxnet.nd import expand_dims [as 别名]
def Route(self, x):
# b_mat = nd.repeat(self.b_mat.data(), repeats=x.shape[0], axis=0)#nd.stop_gradient(nd.repeat(self.b_mat.data(), repeats=x.shape[0], axis=0))
b_mat = nd.zeros((x.shape[0],1,self.num_cap, self.num_locations), ctx=x.context)
x_expand = nd.expand_dims(nd.expand_dims(x, axis=2),2)
w_expand = nd.repeat(nd.expand_dims(self.w_ij.data(x.context),axis=0), repeats=x.shape[0], axis=0)
u_ = w_expand*x_expand
# u_ = nd.abs(w_expand - x_expand)
u = nd.sum(u_, axis = 1)
u_no_gradient = nd.stop_gradient(u)
for i in range(self.route_num):
c_mat = nd.softmax(b_mat, axis=2)
if i == self.route_num -1:
s = nd.sum(u * c_mat, axis=-1)
else:
s = nd.sum(u_no_gradient * c_mat, axis=-1)
v = squash(s, 1)
v1 = nd.expand_dims(v, axis=-1)
if i != self.route_num - 1:
update_term = nd.sum(u_no_gradient*v1, axis=1, keepdims=True)
b_mat = b_mat + update_term
return v
示例3: __init__
# 需要导入模块: from mxnet import nd [as 别名]
# 或者: from mxnet.nd import expand_dims [as 别名]
def __init__(self, dist, edge, hidden_size, prefix=None):
super(Graph, self).__init__(prefix=prefix)
self.dist = dist
self.edge = edge
self.hidden_size = hidden_size
# create graph
self.num_nodes = n = self.dist.shape[0]
src, dst, dist = [], [], []
for i in range(n):
for j in edge[i]:
src.append(j)
dst.append(i)
dist.append(self.dist[j, i])
self.src = src
self.dst = dst
self.dist = mx.nd.expand_dims(mx.nd.array(dist), axis=1)
self.ctx = []
self.graph_on_ctx = []
self.init_model()
示例4: extract_rank_embedding_nd
# 需要导入模块: from mxnet import nd [as 别名]
# 或者: from mxnet.nd import expand_dims [as 别名]
def extract_rank_embedding_nd(rank_dim, feat_dim, wave_length=1000):
rank_range = nd.arange(0, rank_dim)
feat_range = nd.arange(0, feat_dim / 2)
dim_mat = nd.broadcast_power(lhs=nd.full((1,), wave_length),
rhs=(2. / feat_dim) * feat_range)
dim_mat = nd.Reshape(dim_mat, shape=(1, -1))
rank_mat = nd.expand_dims(rank_range, axis=1)
div_mat = nd.broadcast_div(lhs=rank_mat, rhs=dim_mat)
sin_mat = nd.sin(data=div_mat)
cos_mat = nd.cos(data=div_mat)
embedding = nd.concat(sin_mat, cos_mat, dim=1)
return embedding
示例5: extract_multi_position_matrix_nd
# 需要导入模块: from mxnet import nd [as 别名]
# 或者: from mxnet.nd import expand_dims [as 别名]
def extract_multi_position_matrix_nd(bbox):
bbox = nd.transpose(bbox, axes=(1, 0, 2))
xmin, ymin, xmax, ymax = nd.split(data=bbox, num_outputs=4, axis=2)
# [num_fg_classes, num_boxes, 1]
bbox_width = xmax - xmin + 1.
bbox_height = ymax - ymin + 1.
center_x = 0.5 * (xmin + xmax)
center_y = 0.5 * (ymin + ymax)
# [num_fg_classes, num_boxes, num_boxes]
delta_x = nd.broadcast_minus(lhs=center_x,
rhs=nd.transpose(center_x, axes=(0, 2, 1)))
delta_x = nd.broadcast_div(delta_x, bbox_width)
delta_x = nd.log(nd.maximum(nd.abs(delta_x), 1e-3))
delta_y = nd.broadcast_minus(lhs=center_y,
rhs=nd.transpose(center_y, axes=(0, 2, 1)))
delta_y = nd.broadcast_div(delta_y, bbox_height)
delta_y = nd.log(nd.maximum(nd.abs(delta_y), 1e-3))
delta_width = nd.broadcast_div(lhs=bbox_width,
rhs=nd.transpose(bbox_width, axes=(0, 2, 1)))
delta_width = nd.log(delta_width)
delta_height = nd.broadcast_div(lhs=bbox_height,
rhs=nd.transpose(bbox_height, axes=(0, 2, 1)))
delta_height = nd.log(delta_height)
concat_list = [delta_x, delta_y, delta_width, delta_height]
for idx, sym in enumerate(concat_list):
concat_list[idx] = nd.expand_dims(sym, axis=3)
position_matrix = nd.concat(*concat_list, dim=3)
return position_matrix
示例6: forward
# 需要导入模块: from mxnet import nd [as 别名]
# 或者: from mxnet.nd import expand_dims [as 别名]
def forward(self, x):
x1 = nd.expand_dims(x, axis=self.axes)
return x1
示例7: forward
# 需要导入模块: from mxnet import nd [as 别名]
# 或者: from mxnet.nd import expand_dims [as 别名]
def forward(self, x):
conv_out = nd.expand_dims(self.cap(x), axis=2)
# conv_out = nd.expand_dims(self.bn(self.cap(x)), axis=2)
conv_out = conv_out.reshape((0,self.num_cap,-1,0,0))
conv_out = squash(conv_out, 2)
# print conv_out.shape
return conv_out
示例8: route
# 需要导入模块: from mxnet import nd [as 别名]
# 或者: from mxnet.nd import expand_dims [as 别名]
def route(self, u):
b_mat = nd.zeros((u.shape[0], self.num_cap_in, self.num_cap, 1, u.shape[4], u.shape[5]), ctx=u.context)
for i in range(self.route_num):
c_mat = nd.softmax(b_mat, axis=2)
s = nd.sum(u * c_mat, axis=1)
v = squash(s, 2)
if i != self.route_num - 1:
v1 = nd.expand_dims(v, axis=1)
update_term = nd.sum(u*v1, axis=3, keepdims=True)
b_mat = b_mat + update_term
return v
示例9: forward
# 需要导入模块: from mxnet import nd [as 别名]
# 或者: from mxnet.nd import expand_dims [as 别名]
def forward(self, x):
conv_out = nd.expand_dims(self.cap(x), axis=2)
conv_out = conv_out.reshape((0,-1,self.num_cap,0,0))
conv_out = squash(conv_out, 1)
return conv_out
示例10: msg_edge
# 需要导入模块: from mxnet import nd [as 别名]
# 或者: from mxnet.nd import expand_dims [as 别名]
def msg_edge(self, edge):
state = nd.concat(edge.src['state'], edge.dst['state'], dim=-1)
ctx = state.context
alpha = nd.LeakyReLU(nd.dot(state, self.weight.data(ctx)))
dist = edge.data['dist']
while len(dist.shape) < len(alpha.shape):
dist = nd.expand_dims(dist, axis=-1)
alpha = alpha * dist
return { 'alpha': alpha, 'state': edge.src['state'] }
示例11: forward_single
# 需要导入模块: from mxnet import nd [as 别名]
# 或者: from mxnet.nd import expand_dims [as 别名]
def forward_single(self, feature, data, begin_state):
# add a temporal axis
data = nd.expand_dims(data, axis=2)
# unroll
data, state = self(feature, data, begin_state)
# remove the temporal axis
data = nd.mean(data, axis=2)
return data, state
示例12: init_weights
# 需要导入模块: from mxnet import nd [as 别名]
# 或者: from mxnet.nd import expand_dims [as 别名]
def init_weights(self, ctx):
"""Initial I3D network with its 2D pretrained weights."""
self.first_stage.initialize(ctx=ctx)
self.res_layers.initialize(ctx=ctx)
self.head.initialize(ctx=ctx)
if self.pretrained_base and not self.pretrained:
if self.depth == 50:
resnet2d = resnet50_v1b(pretrained=True)
elif self.depth == 101:
resnet2d = resnet101_v1b(pretrained=True)
else:
print('No such 2D pre-trained network of depth %d.' % (self.depth))
weights2d = resnet2d.collect_params()
if self.nonlocal_cfg is None:
weights3d = self.collect_params()
else:
train_params_list = []
raw_params = self.collect_params()
for raw_name in raw_params.keys():
if 'nonlocal' in raw_name:
continue
train_params_list.append(raw_name)
init_patterns = '|'.join(train_params_list)
weights3d = self.collect_params(init_patterns)
assert len(weights2d.keys()) == len(weights3d.keys()), 'Number of parameters should be same.'
dict2d = {}
for key_id, key_name in enumerate(weights2d.keys()):
dict2d[key_id] = key_name
dict3d = {}
for key_id, key_name in enumerate(weights3d.keys()):
dict3d[key_id] = key_name
dict_transform = {}
for key_id, key_name in dict3d.items():
dict_transform[dict2d[key_id]] = key_name
cnt = 0
for key2d, key3d in dict_transform.items():
if 'conv' in key3d:
temporal_dim = weights3d[key3d].shape[2]
temporal_2d = nd.expand_dims(weights2d[key2d].data(), axis=2)
inflated_2d = nd.broadcast_to(temporal_2d, shape=[0, 0, temporal_dim, 0, 0]) / temporal_dim
assert inflated_2d.shape == weights3d[key3d].shape, 'the shape of %s and %s does not match. ' % (key2d, key3d)
weights3d[key3d].set_data(inflated_2d)
cnt += 1
print('%s is done with shape: ' % (key3d), weights3d[key3d].shape)
if 'batchnorm' in key3d:
assert weights2d[key2d].shape == weights3d[key3d].shape, 'the shape of %s and %s does not match. ' % (key2d, key3d)
weights3d[key3d].set_data(weights2d[key2d].data())
cnt += 1
print('%s is done with shape: ' % (key3d), weights3d[key3d].shape)
if 'dense' in key3d:
cnt += 1
print('%s is skipped with shape: ' % (key3d), weights3d[key3d].shape)
assert cnt == len(weights2d.keys()), 'Not all parameters have been ported, check the initialization.'
示例13: refine_bbox_nd
# 需要导入模块: from mxnet import nd [as 别名]
# 或者: from mxnet.nd import expand_dims [as 别名]
def refine_bbox_nd(bbox, bbox_delta, im_info=None, means=None, stds=None):
xmin, ymin, xmax, ymax = nd.split(data=bbox, num_outputs=4, axis=1)
bbox_width = xmax - xmin + 1.
bbox_height = ymax - ymin + 1.
center_x = 0.5 * (xmin + xmax)
center_y = 0.5 * (ymin + ymax)
bbox_delta_reshape = nd.Reshape(data=bbox_delta, shape=(0, -1, 4))
dx, dy, dw, dh = nd.split(data=bbox_delta_reshape,
num_outputs=4, axis=2, squeeze_axis=1)
if (means is not None) and (stds is not None):
dx = dx * stds[0] + means[0]
dy = dy * stds[1] + means[1]
dw = dw * stds[2] + means[2]
dh = dh * stds[3] + means[3]
refine_center_x = nd.broadcast_add(lhs=center_x,
rhs=nd.broadcast_mul(lhs=bbox_width, rhs=dx))
refine_center_y = nd.broadcast_add(lhs=center_y,
rhs=nd.broadcast_mul(lhs=bbox_height, rhs=dy))
refined_width = nd.broadcast_mul(lhs=bbox_width, rhs=nd.exp(dw))
refined_height = nd.broadcast_mul(lhs=bbox_height, rhs=nd.exp(dh))
w_offset = 0.5 * (refined_width - 1.)
h_offset = 0.5 * (refined_height - 1.)
refined_xmin = nd.expand_dims(refine_center_x - w_offset, axis=1)
refined_ymin = nd.expand_dims(refine_center_y - h_offset, axis=1)
refined_xmax = nd.expand_dims(refine_center_x + w_offset, axis=1)
refined_ymax = nd.expand_dims(refine_center_y + h_offset, axis=1)
refined_bbox = nd.concat(refined_xmin, refined_ymin, refined_xmax, refined_ymax, dim=1)
if im_info is not None:
# assume im_info [[height, width, scale]] with shape (1,3)
im_hw = nd.slice_axis(im_info, axis=1, begin=0, end=2)
im_wh = nd.reverse(im_hw, axis=1)
im_wh = im_wh - 1.
im_wh = nd.tile(data=im_wh, reps=(1, 2))
im_wh = nd.Reshape(im_wh, shape=(1, 4, 1))
refined_bbox = nd.broadcast_minimum(lhs=refined_bbox, rhs=im_wh)
refined_bbox = nd.broadcast_maximum(lhs=refined_bbox,
rhs=nd.zeros_like(refined_bbox))
# print refined_bbox.debug_str()
return refined_bbox
示例14: forward
# 需要导入模块: from mxnet import nd [as 别名]
# 或者: from mxnet.nd import expand_dims [as 别名]
def forward(self, feature, label, begin_states, is_training):
''' Decode the hidden states to a temporal sequence.
Parameters
----------
feature: a NDArray with shape [n, d].
label: a NDArray with shape [n, b, t, d].
begin_states: a list of hidden states (list of hidden units with shape [n, b, d]) of RNNs.
is_training: bool
Returns
-------
outputs: the prediction, which is a NDArray with shape [n, b, t, d]
'''
ctx = label.context
num_nodes, batch_size, seq_len, _ = label.shape
aux = label[:,:,:, self.output_dim:] # [n,b,t,d]
label = label[:,:,:, :self.output_dim] # [n,b,t,d]
go = nd.zeros(shape=(num_nodes, batch_size, self.input_dim), ctx=ctx)
output, states = [], begin_states
for i in range(seq_len):
# get next input
if i == 0: data = go
else:
prev = nd.concat(output[i - 1], aux[:,:,i - 1], dim=-1)
truth = nd.concat(label[:,:,i - 1], aux[:,:,i - 1], dim=-1)
if is_training and self.use_sampling: value = self.sampling()
else: value = 0
data = value * truth + (1 - value) * prev
# unroll 1 step
for depth, cell in enumerate(self.cells):
data, states[depth] = cell.forward_single(feature, data, states[depth])
if self.graphs[depth] is not None:
_data = 0
for g in self.graphs[depth]:
_data = _data + g(data, feature)
data = _data / len(self.graphs[depth])
# append feature to output
_feature = nd.expand_dims(feature, axis=1) # [n, 1, d]
_feature = nd.broadcast_to(_feature, shape=(0, batch_size, 0)) # [n, b, d]
data = nd.concat(data, _feature, dim=-1) # [n, b, t, d]
# proj output to prediction
data = nd.reshape(data, shape=(num_nodes * batch_size, -1))
data = self.proj(data)
data = nd.reshape(data, shape=(num_nodes, batch_size, -1))
output.append(data)
output = nd.stack(*output, axis=2)
return output
示例15: forward
# 需要导入模块: from mxnet import nd [as 别名]
# 或者: from mxnet.nd import expand_dims [as 别名]
def forward(self, feature, label, begin_states, is_training):
''' Decode the hidden states to a temporal sequence.
Parameters
----------
feature: a NDArray with shape [n, d].
label: a NDArray with shape [n, b, t, d].
begin_states: a list of hidden states (list of hidden units with shape [n, b, d]) of RNNs.
is_training: bool
Returns
-------
outputs: the prediction, which is a NDArray with shape [n, b, t, d]
'''
ctx = label.context
num_nodes, batch_size, seq_len, _ = label.shape
aux = label[:,:,:, self.output_dim:] # [n,b,t,d]
label = label[:,:,:, :self.output_dim] # [n,b,t,d]
go = nd.zeros(shape=(num_nodes, batch_size, self.input_dim), ctx=ctx)
output, states = [], begin_states
for i in range(seq_len):
# get next input
if i == 0: data = go
else:
prev = nd.concat(output[i - 1], aux[:,:,i - 1], dim=-1)
truth = nd.concat(label[:,:,i - 1], aux[:,:,i - 1], dim=-1)
if is_training and self.use_sampling: value = self.sampling()
else: value = 0
data = value * truth + (1 - value) * prev
# unroll 1 step
for depth, cell in enumerate(self.cells):
data, states[depth] = cell.forward_single(feature, data, states[depth])
if self.graphs[depth] is not None:
_data = data
for g in self.graphs[depth]:
_data = _data + g(data, feature)
data = _data
# append feature to output
_feature = nd.expand_dims(feature, axis=1) # [n, 1, d]
_feature = nd.broadcast_to(_feature, shape=(0, batch_size, 0)) # [n, b, d]
data = nd.concat(data, _feature, dim=-1) # [n, b, t, d]
# proj output to prediction
data = nd.reshape(data, shape=(num_nodes * batch_size, -1))
data = self.proj(data)
data = nd.reshape(data, shape=(num_nodes, batch_size, -1))
output.append(data)
output = nd.stack(*output, axis=2)
return output