-
Notifications
You must be signed in to change notification settings - Fork 22
/
Copy pathWarpST.py
144 lines (125 loc) · 5.77 KB
/
WarpST.py
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
import tensorflow as tf
from bicubic_interp import bicubic_interp_2d
def WarpST(U, V, out_size, name='DeformableTransformer', **kwargs):
"""Deformable Transformer Layer with bicubic interpolation
U : tf.float, [num_batch, height, width, num_channels].
Input tensor to warp
V : tf.float, [num_batch, height, width, 2]
Warp map. It is interpolated to out_size.
out_size: a tuple of two ints
The size of the output of the network (height, width)
----------
References :
https://github.com/daviddao/spatial-transformer-tensorflow/blob/master/spatial_transformer.py
"""
def _repeat(x, n_repeats):
with tf.variable_scope('_repeat'):
rep = tf.transpose(
tf.expand_dims(tf.ones(shape=tf.stack([n_repeats, ])), 1), [1, 0])
rep = tf.cast(rep, 'int32')
x = tf.matmul(tf.reshape(x, (-1, 1)), rep)
return tf.reshape(x, [-1])
def _interpolate(im, x, y, out_size):
with tf.variable_scope('_interpolate'):
# constants
num_batch = tf.shape(im)[0]
height = tf.shape(im)[1]
width = tf.shape(im)[2]
channels = tf.shape(im)[3]
x = tf.cast(x, 'float32')
y = tf.cast(y, 'float32')
height_f = tf.cast(height, 'float32')
width_f = tf.cast(width, 'float32')
out_height = out_size[0]
out_width = out_size[1]
zero = tf.zeros([], dtype='int32')
max_y = tf.cast(tf.shape(im)[1] - 1, 'int32')
max_x = tf.cast(tf.shape(im)[2] - 1, 'int32')
# scale indices from [-1, 1] to [0, width/height]
x = (x + 1.0)*(width_f) / 2.0
y = (y + 1.0)*(height_f) / 2.0
# do sampling
x0 = tf.cast(tf.floor(x), 'int32')
x1 = x0 + 1
y0 = tf.cast(tf.floor(y), 'int32')
y1 = y0 + 1
x0 = tf.clip_by_value(x0, zero, max_x)
x1 = tf.clip_by_value(x1, zero, max_x)
y0 = tf.clip_by_value(y0, zero, max_y)
y1 = tf.clip_by_value(y1, zero, max_y)
dim2 = width
dim1 = width*height
base = _repeat(tf.range(num_batch)*dim1, out_height*out_width)
base_y0 = base + y0*dim2
base_y1 = base + y1*dim2
idx_a = base_y0 + x0
idx_b = base_y1 + x0
idx_c = base_y0 + x1
idx_d = base_y1 + x1
# use indices to lookup pixels in the flat image and restore
# channels dim
im_flat = tf.reshape(im, tf.stack([-1, channels]))
im_flat = tf.cast(im_flat, 'float32')
Ia = tf.gather(im_flat, idx_a)
Ib = tf.gather(im_flat, idx_b)
Ic = tf.gather(im_flat, idx_c)
Id = tf.gather(im_flat, idx_d)
# and finally calculate interpolated values
x0_f = tf.cast(x0, 'float32')
x1_f = tf.cast(x1, 'float32')
y0_f = tf.cast(y0, 'float32')
y1_f = tf.cast(y1, 'float32')
wa = tf.expand_dims(((x1_f-x) * (y1_f-y)), 1)
wb = tf.expand_dims(((x1_f-x) * (y-y0_f)), 1)
wc = tf.expand_dims(((x-x0_f) * (y1_f-y)), 1)
wd = tf.expand_dims(((x-x0_f) * (y-y0_f)), 1)
output = tf.add_n([wa*Ia, wb*Ib, wc*Ic, wd*Id])
return output
def _meshgrid(height, width):
with tf.variable_scope('_meshgrid'):
# This should be equivalent to:
# x_t, y_t = np.meshgrid(np.linspace(-1, 1, width),
# np.linspace(-1, 1, height))
# ones = np.ones(np.prod(x_t.shape))
# grid = np.vstack([x_t.flatten(), y_t.flatten(), ones])
x_t = tf.matmul(tf.ones(shape=tf.stack([height, 1])),
tf.transpose(tf.expand_dims(tf.linspace(-1.0, 1.0, width), 1), [1, 0]))
y_t = tf.matmul(tf.expand_dims(tf.linspace(-1.0, 1.0, height), 1),
tf.ones(shape=tf.stack([1, width])))
x_t_flat = tf.reshape(x_t, (1, -1))
y_t_flat = tf.reshape(y_t, (1, -1))
grid = tf.concat([x_t_flat, y_t_flat], 0)
return grid
def _transform(V, U, out_size):
with tf.variable_scope('_transform'):
num_batch = tf.shape(U)[0]
height = tf.shape(U)[1]
width = tf.shape(U)[2]
num_channels = tf.shape(U)[3]
# grid of (x_t, y_t, 1), eq (1) in ref [1]
height_f = tf.cast(height, 'float32')
width_f = tf.cast(width, 'float32')
out_height = out_size[0]
out_width = out_size[1]
grid = _meshgrid(out_height, out_width) # [2, h*w]
grid = tf.reshape(grid, [-1]) # [2*h*w]
grid = tf.tile(grid, tf.stack([num_batch])) # [n*2*h*w]
grid = tf.reshape(grid, tf.stack([num_batch, 2, -1])) # [n, 2, h*w]
# Set source position (x+vx, y+vy)^T
V = bicubic_interp_2d(V, out_size)
V = tf.transpose(V, [0, 3, 1, 2]) # [n, 2, h, w]
V = tf.reshape(V, [num_batch, 2, -1]) # [n, 2, h*w]
T_g = tf.add(V, grid) # [n, 2, h*w]
x_s = tf.slice(T_g, [0, 0, 0], [-1, 1, -1])
y_s = tf.slice(T_g, [0, 1, 0], [-1, 1, -1])
x_s_flat = tf.reshape(x_s, [-1])
y_s_flat = tf.reshape(y_s, [-1])
input_transformed = _interpolate(
U, x_s_flat, y_s_flat, out_size)
output = tf.reshape(
input_transformed,
tf.stack([num_batch, out_height, out_width, num_channels]))
return output
with tf.variable_scope(name):
output = _transform(V, U, out_size)
return output