-
Notifications
You must be signed in to change notification settings - Fork 23
/
eval_coco.py
198 lines (164 loc) · 6.85 KB
/
eval_coco.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
# Originated from https://github.com/amdegroot/ssd.pytorch/issues/422
"""Adapted from:
@longcw faster_rcnn_pytorch: https://github.com/longcw/faster_rcnn_pytorch
@rbgirshick py-faster-rcnn https://github.com/rbgirshick/py-faster-rcnn
Licensed under The MIT License
"""
import torch
from torch.autograd import Variable
from data import COCODetection_eval, COCO_ROOT_EVAL
from data import COCO_CLASSES as labelmap
import torch.nn as nn
from ssd_gmm import build_ssd_gmm
import cv2
import os
import time
import pickle
import argparse
import numpy as np
def str2bool(v):
return v.lower() in ("yes", "true", "t", "1")
parser = argparse.ArgumentParser(
description='Single Shot MultiBox Detector Evaluation')
parser.add_argument('--dataset_root',
default=COCO_ROOT_EVAL, type=str,
help='path to your coco2017 data')
parser.add_argument('--model_type',
default='SSD', type=str,
help='tested model')
parser.add_argument('--trained_model',
default='weights/trained_COCO_name.pth', type=str,
help='Trained state_dict file path to open')
parser.add_argument('--save_folder', default='eval/', type=str,
help='File path to save results')
parser.add_argument('--cuda', default=True, type=str2bool,
help='Use cuda to train model')
parser.add_argument('--cleanup', default=True, type=str2bool,
help='Cleanup and remove results files following eval')
parser.add_argument('--retest', default=False, type=str2bool,
help='test the result on result file')
args = parser.parse_args()
if not os.path.exists(args.save_folder):
os.makedirs(args.save_folder)
if torch.cuda.is_available():
if args.cuda:
torch.set_default_tensor_type('torch.cuda.FloatTensor')
if not args.cuda:
print("WARNING: It looks like you have a CUDA device, but aren't using \
CUDA. Run with --cuda for optimal eval speed.")
torch.set_default_tensor_type('torch.FloatTensor')
else:
torch.set_default_tensor_type('torch.FloatTensor')
class Timer(object):
"""A simple timer."""
def __init__(self):
self.total_time = 0.
self.calls = 0
self.start_time = 0.
self.diff = 0.
self.average_time = 0.
def tic(self):
# using time.time instead of time.clock because time time.clock
# does not normalize for multithreading
self.start_time = time.time()
def toc(self, average=True):
self.diff = time.time() - self.start_time
self.total_time += self.diff
self.calls += 1
self.average_time = self.total_time / self.calls
if average:
return self.average_time
else:
return self.diff
class BaseTransform(object):
"""Defines the transformations that should be applied to test PIL image
for input into the network
dimension -> tensorize -> color adj
Arguments:
resize (int): input dimension to SSD
rgb_means ((int,int,int)): average RGB of the dataset
(104,117,123)
rgb_std: std of the dataset
swap ((int,int,int)): final order of channels
Returns:
transform (transform) : callable transform to be applied to test/val
data
"""
def __init__(self, resize, rgb_means, rgb_std=(1, 1, 1), swap=(2, 0, 1)):
self.means = rgb_means
self.resize = resize
self.std = rgb_std
self.swap = swap
# assume input is cv2 img for now
def __call__(self, img):
img = cv2.resize(np.array(img), (self.resize, self.resize)).astype(np.float32)
img -= self.means
img /= self.std
img = img.transpose(self.swap)
return torch.from_numpy(img)
def test_net(save_folder, net, cuda, testset, transform):
with torch.no_grad():
if not os.path.exists(save_folder):
os.makedirs(save_folder)
# dump predictions and assoc. ground truth to text file for now
num_images = len(testset)
num_classes = 81
all_boxes = [[[] for _ in range(num_images)]
for _ in range(num_classes)]
_t = {'im_detect': Timer(), 'misc': Timer()}
det_file = os.path.join(save_folder, 'detections.pkl')
if args.retest:
f = open(det_file, 'rb')
all_boxes = pickle.load(f)
print('Evaluating detections')
testset.evaluate_detections(all_boxes, save_folder)
return
for i in range(num_images):
img, h, w = testset.pull_image(i)
x = Variable(transform(img).unsqueeze(0))
if cuda:
x = x.cuda()
_t['im_detect'].tic()
detections = net(x).data # [1, class, top_k, 5]
detect_time = _t['im_detect'].toc(average=False)
# skip j = 0, because it's the background class
for j in range(1, detections.size(1)):
dets = detections[0, j, :, :] # [top_k, 5]
mask = dets[:, 0].gt(0.).expand(15, dets.size(0)).t()
dets = torch.masked_select(dets, mask).view(-1, 15)
if dets.size(0) == 0:
continue
boxes = dets[:, 1:5]
boxes[:, 0] *= w
boxes[:, 2] *= w
boxes[:, 1] *= h
boxes[:, 3] *= h
scores = dets[:, 0].cpu().numpy()
cls_dets = np.hstack((boxes.cpu().numpy(),
scores[:, np.newaxis])).astype(np.float32, copy=False)
all_boxes[j][i] = cls_dets
print('im_detect: {:d}/{:d} {:.3f}s'.format(i + 1, num_images, detect_time), end='\r')
with open(det_file, 'wb') as f:
pickle.dump(all_boxes, f, pickle.HIGHEST_PROTOCOL)
print('Evaluating detections')
testset.evaluate_detections(all_boxes, save_folder)
if __name__ == '__main__':
# load net
num_classes = len(labelmap) + 1 # +1 for background
print(num_classes)
net = build_ssd_gmm('test', size=300, num_classes=num_classes)
net = nn.DataParallel(net)
print(torch.cuda.is_available())
if args.trained_model:
print("Loading weight:", args.trained_model)
ckp = torch.load(args.trained_model)
net.load_state_dict(ckp['weight'] if 'weight' in ckp.keys() else ckp)
net.eval()
print('Finished loading model!')
# test on coco2017 VAL set (5000 images)
testset = COCODetection_eval(args.dataset_root, [('2017', 'val')], None)
# save the test result here (those detected bounding box, etc.)
save_folder = os.path.join(args.save_folder, 'coco')
test_save_dir = os.path.join(save_folder, args.model_type)
test_net(test_save_dir, net, args.cuda, testset,
BaseTransform(300, rgb_means=(123, 117, 104), rgb_std=(1, 1, 1), swap=(2, 0, 1)))