-
Notifications
You must be signed in to change notification settings - Fork 41
/
evaluators.py
201 lines (171 loc) · 7.32 KB
/
evaluators.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
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
from __future__ import print_function, absolute_import
import time
from collections import OrderedDict
import numpy as np
from sklearn.preprocessing import normalize
from sklearn.metrics import pairwise_distances
import torch
import torch.nn.functional as F
import torch.distributed as dist
import torch.multiprocessing as mp
from torch.utils.data import DataLoader
from .pca import PCA
from .utils.meters import AverageMeter
from .utils.rerank import re_ranking
from .utils.dist_utils import synchronize
from .utils.serialization import write_json
from .utils.data.preprocessor import Preprocessor
from .utils import to_torch
def extract_cnn_feature(model, inputs, vlad=True, gpu=None):
model.eval()
inputs = to_torch(inputs).cuda(gpu)
outputs = model(inputs)
if (isinstance(outputs, list) or isinstance(outputs, tuple)):
x_pool, x_vlad = outputs
if vlad:
outputs = F.normalize(x_vlad, p=2, dim=-1)
else:
outputs = F.normalize(x_pool, p=2, dim=-1)
else:
outputs = F.normalize(outputs, p=2, dim=-1)
return outputs
def extract_features(model, data_loader, dataset, print_freq=10,
vlad=True, pca=None, gpu=None, sync_gather=False):
model.eval()
batch_time = AverageMeter()
data_time = AverageMeter()
rank = dist.get_rank()
world_size = dist.get_world_size()
features = []
if (pca is not None):
pca.load(gpu=gpu)
end = time.time()
with torch.no_grad():
for i, (imgs, fnames, _, _, _) in enumerate(data_loader):
data_time.update(time.time() - end)
outputs = extract_cnn_feature(model, imgs, vlad, gpu=gpu)
if (pca is not None):
outputs = pca.infer(outputs)
outputs = outputs.data.cpu()
features.append(outputs)
batch_time.update(time.time() - end)
end = time.time()
if ((i + 1) % print_freq == 0 and rank==0):
print('Extract Features: [{}/{}]\t'
'Time {:.3f} ({:.3f})\t'
'Data {:.3f} ({:.3f})\t'
.format(i + 1, len(data_loader),
batch_time.val, batch_time.avg,
data_time.val, data_time.avg))
if (pca is not None):
del pca
if (sync_gather):
# all gather features in parallel
# cost more GPU memory but less time
features = torch.cat(features).cuda(gpu)
all_features = [torch.empty_like(features) for _ in range(world_size)]
dist.all_gather(all_features, features)
del features
all_features = torch.cat(all_features).cpu()[:len(dataset)]
features_dict = OrderedDict()
for fname, output in zip(dataset, all_features):
features_dict[fname[0]] = output
del all_features
else:
# broadcast features in sequence
# cost more time but less GPU memory
bc_features = torch.cat(features).cuda(gpu)
features_dict = OrderedDict()
for k in range(world_size):
bc_features.data.copy_(torch.cat(features))
if (rank==0):
print("gathering features from rank no.{}".format(k))
dist.broadcast(bc_features, k)
l = bc_features.cpu().size(0)
for fname, output in zip(dataset[k*l:(k+1)*l], bc_features.cpu()):
features_dict[fname[0]] = output
del bc_features, features
return features_dict
def pairwise_distance(features, query=None, gallery=None, metric=None):
if query is None and gallery is None:
n = len(features)
x = torch.cat(list(features.values()))
x = x.view(n, -1)
if metric is not None:
x = metric.transform(x)
dist_m = torch.pow(x, 2).sum(dim=1, keepdim=True) * 2
dist_m = dist_m.expand(n, n) - 2 * torch.mm(x, x.t())
return dist_m, None, None
if (dist.get_rank()==0):
print ("===> Start calculating pairwise distances")
x = torch.cat([features[f].unsqueeze(0) for f, _, _, _ in query], 0)
y = torch.cat([features[f].unsqueeze(0) for f, _, _, _ in gallery], 0)
m, n = x.size(0), y.size(0)
x = x.view(m, -1)
y = y.view(n, -1)
if metric is not None:
x = metric.transform(x)
y = metric.transform(y)
dist_m = torch.pow(x, 2).sum(dim=1, keepdim=True).expand(m, n) + \
torch.pow(y, 2).sum(dim=1, keepdim=True).expand(n, m).t()
dist_m.addmm_(1, -2, x, y.t())
return dist_m, x.numpy(), y.numpy()
def spatial_nms(pred, db_ids, topN):
assert(len(pred)==len(db_ids))
pred_select = pred[:topN]
pred_pids = [db_ids[i] for i in pred_select]
# find unique
seen = set()
seen_add = seen.add
pred_pids_unique = [i for i, x in enumerate(pred_pids) if not (x in seen or seen_add(x))]
return [pred_select[i] for i in pred_pids_unique]
def evaluate_all(distmat, gt, gallery, recall_topk=[1, 5, 10], nms=False):
sort_idx = np.argsort(distmat, axis=1)
del distmat
db_ids = [db[1] for db in gallery]
if (dist.get_rank()==0):
print("===> Start calculating recalls")
correct_at_n = np.zeros(len(recall_topk))
for qIx, pred in enumerate(sort_idx):
if (nms):
pred = spatial_nms(pred.tolist(), db_ids, max(recall_topk)*12)
for i, n in enumerate(recall_topk):
# if in top N then also in top NN, where NN > N
if np.any(np.in1d(pred[:n], gt[qIx])):
correct_at_n[i:] += 1
break
recalls = correct_at_n / len(gt)
del sort_idx
if (dist.get_rank()==0):
print('Recall Scores:')
for i, k in enumerate(recall_topk):
print(' top-{:<4}{:12.1%}'.format(k, recalls[i]))
return recalls
class Evaluator(object):
def __init__(self, model):
super(Evaluator, self).__init__()
self.model = model
self.rank = dist.get_rank()
def evaluate(self, query_loader, dataset, query, gallery, ground_truth, gallery_loader=None, \
vlad=True, pca=None, rerank=False, gpu=None, sync_gather=False, \
nms=False, rr_topk=25, lambda_value=0):
if (gallery_loader is not None):
features = extract_features(self.model, query_loader, query,
vlad=vlad, pca=pca, gpu=gpu, sync_gather=sync_gather)
features_db = extract_features(self.model, gallery_loader, gallery,
vlad=vlad, pca=pca, gpu=gpu, sync_gather=sync_gather)
features.update(features_db)
else:
features = extract_features(self.model, query_loader, dataset,
vlad=vlad, pca=pca, gpu=gpu, sync_gather=sync_gather)
distmat, _, _ = pairwise_distance(features, query, gallery)
recalls = evaluate_all(distmat, ground_truth, gallery, nms=nms)
if (not rerank):
return recalls
if (self.rank==0):
print('Applying re-ranking ...')
distmat_gg, _, _ = pairwise_distance(features, gallery, gallery)
distmat_qq, _, _ = pairwise_distance(features, query, query)
distmat = re_ranking(distmat.numpy(), distmat_qq.numpy(), distmat_gg.numpy(),
k1=rr_topk, k2=1, lambda_value=lambda_value)
return evaluate_all(distmat, ground_truth, gallery, nms=nms)