-
Notifications
You must be signed in to change notification settings - Fork 3
/
evaluation.py
172 lines (144 loc) · 5.72 KB
/
evaluation.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
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
import numpy as np
#import numpy
def i2t_SCAN(sims, npts=None, return_ranks=False):
"""
Images->Text (Image Annotation)
Images: (N, n_region, d) matrix of images
Captions: (5N, max_n_word, d) matrix of captions
CapLens: (5N) array of caption lengths
sims: (N, 5N) matrix of similarity im-cap
"""
npts = sims.shape[0]
ranks = np.zeros(npts)
top1 = np.zeros(npts)
for index in range(npts):
inds = np.argsort(sims[index])[::-1]
# Score
rank = 1e20
for i in range(5 * index, 5 * index + 5, 1):
tmp = np.where(inds == i)[0][0]
if tmp < rank:
rank = tmp
ranks[index] = rank
top1[index] = inds[0]
# Compute metrics
r1 = 100.0 * len(np.where(ranks < 1)[0]) / len(ranks)
r5 = 100.0 * len(np.where(ranks < 5)[0]) / len(ranks)
r10 = 100.0 * len(np.where(ranks < 10)[0]) / len(ranks)
r20 = 100.0 * len(np.where(ranks < 20)[0]) / len(ranks)
r50 = 100.0 * len(np.where(ranks < 50)[0]) / len(ranks)
r70 = 100.0 * len(np.where(ranks < 70)[0]) / len(ranks)
r100 = 100.0 * len(np.where(ranks < 100)[0]) / len(ranks)
medr = np.floor(np.median(ranks)) + 1
meanr = ranks.mean() + 1
if return_ranks:
return (r1, r5, r10, r20, r50, r70, r100, medr, meanr), (ranks, top1)
else:
return (r1, r5, r10, r20, r50, r70, r100, medr, meanr)
def t2i_SCAN(sims, npts=None, return_ranks=False):
"""
Text->Images (Image Search)
Images: (N, n_region, d) matrix of images
Captions: (5N, max_n_word, d) matrix of captions
CapLens: (5N) array of caption lengths
sims: (N, 5N) matrix of similarity im-cap
"""
npts = sims.shape[0]
ranks = np.zeros(5 * npts)
top1 = np.zeros(5 * npts)
# --> (5N(caption), N(image))
sims = sims.T
for index in range(npts):
for i in range(5):
inds = np.argsort(sims[5 * index + i])[::-1]
ranks[5 * index + i] = np.where(inds == index)[0][0]
top1[5 * index + i] = inds[0]
# Compute metrics
r1 = 100.0 * len(np.where(ranks < 1)[0]) / len(ranks)
r5 = 100.0 * len(np.where(ranks < 5)[0]) / len(ranks)
r10 = 100.0 * len(np.where(ranks < 10)[0]) / len(ranks)
r20 = 100.0 * len(np.where(ranks < 20)[0]) / len(ranks)
r50 = 100.0 * len(np.where(ranks < 50)[0]) / len(ranks)
r70 = 100.0 * len(np.where(ranks < 70)[0]) / len(ranks)
r100 = 100.0 * len(np.where(ranks < 100)[0]) / len(ranks)
medr = np.floor(np.median(ranks)) + 1
meanr = ranks.mean() + 1
if return_ranks:
return (r1, r5, r10, r20, r50, r70, r100, medr, meanr), (ranks, top1)
else:
return (r1, r5, r10, r20, r50, r70, r100, medr, meanr)
'''def i2t(images, captions, npts=None, measure='cosine', return_ranks=False):
"""
Images->Text (Image Annotation)
Images: (5N, K) matrix of images
Captions: (5N, K) matrix of captions
"""
if npts is None:
npts = int(images.shape[0] / 5)
index_list = []
ranks = numpy.zeros(npts)
top1 = numpy.zeros(npts)
for index in range(npts):
# Get query image
im = images[5 * index].reshape(1, images.shape[1])
# Compute scores
d = numpy.dot(im, captions.T).flatten()
inds = numpy.argsort(d)[::-1]
index_list.append(inds[0])
# Score
rank = 1e20
for i in range(5 * index, 5 * index + 5, 1):
tmp = numpy.where(inds == i)[0][0]
if tmp < rank:
rank = tmp
ranks[index] = rank
top1[index] = inds[0]
# Compute metrics
r1 = 100.0 * len(numpy.where(ranks < 1)[0]) / len(ranks)
r5 = 100.0 * len(numpy.where(ranks < 5)[0]) / len(ranks)
r10 = 100.0 * len(numpy.where(ranks < 10)[0]) / len(ranks)
r20 = 100.0 * len(numpy.where(ranks < 20)[0]) / len(ranks)
r50 = 100.0 * len(numpy.where(ranks < 50)[0]) / len(ranks)
r70 = 100.0 * len(numpy.where(ranks < 70)[0]) / len(ranks)
r100 = 100.0 * len(numpy.where(ranks < 100)[0]) / len(ranks)
medr = numpy.floor(numpy.median(ranks)) + 1
meanr = ranks.mean() + 1
if return_ranks:
return (r1, r5, r10, r20, r50, r70, r100, medr, meanr), (ranks, top1)
else:
return (r1, r5, r10, r20, r50, r70, r100, medr, meanr)
def t2i(images, captions, npts=None, measure='cosine', return_ranks=False):
"""
Text->Images (Image Search)
Images: (5N, K) matrix of images
Captions: (5N, K) matrix of captions
"""
if npts is None:
npts = int(images.shape[0] / 5)
ims = numpy.array([images[i] for i in range(0, len(images), 5)])
ranks = numpy.zeros(5 * npts)
top1 = numpy.zeros(5 * npts)
for index in range(npts):
# Get query captions
queries = captions[5 * index:5 * index + 5]
# Compute scores
d = numpy.dot(queries, ims.T)
inds = numpy.zeros(d.shape)
for i in range(len(inds)):
inds[i] = numpy.argsort(d[i])[::-1]
ranks[5 * index + i] = numpy.where(inds[i] == index)[0][0]
top1[5 * index + i] = inds[i][0]
# Compute metrics
r1 = 100.0 * len(numpy.where(ranks < 1)[0]) / len(ranks)
r5 = 100.0 * len(numpy.where(ranks < 5)[0]) / len(ranks)
r10 = 100.0 * len(numpy.where(ranks < 10)[0]) / len(ranks)
r20 = 100.0 * len(numpy.where(ranks < 20)[0]) / len(ranks)
r50 = 100.0 * len(numpy.where(ranks < 50)[0]) / len(ranks)
r70 = 100.0 * len(numpy.where(ranks < 70)[0]) / len(ranks)
r100 = 100.0 * len(numpy.where(ranks < 100)[0]) / len(ranks)
medr = numpy.floor(numpy.median(ranks)) + 1
meanr = ranks.mean() + 1
if return_ranks:
return (r1, r5, r10, r20, r50, r70, r100, medr, meanr), (ranks, top1)
else:
return (r1, r5, r10, r20, r50, r70, r100, medr, meanr)'''