-
Notifications
You must be signed in to change notification settings - Fork 1
/
Copy pathtest.py
122 lines (106 loc) · 6.18 KB
/
test.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
from absl import app, flags, logging
from absl.flags import FLAGS
import cv2
import os
import numpy as np
import tensorflow as tf
import csv
from modules.evaluations import get_val_data, perform_val,perform_val_yts
from modules.models import ArcFaceModel
from modules.utils import set_memory_growth, load_yaml, l2_norm
flags.DEFINE_string('cfg_path', './configs/iom_res50.yaml', 'config file path')
flags.DEFINE_string('gpu', '0', 'which gpu to use')
flags.DEFINE_string('img_path', '', 'path to input image')
def main(_argv):
os.environ['TF_CPP_MIN_LOG_LEVEL'] = '3'
os.environ['CUDA_VISIBLE_DEVICES'] = FLAGS.gpu
logger = tf.get_logger()
logger.disabled = True
logger.setLevel(logging.FATAL)
set_memory_growth()
cfg = load_yaml(FLAGS.cfg_path)
model = ArcFaceModel(size=cfg['input_size'],
embd_shape=cfg['embd_shape'],
backbone_type=cfg['backbone_type'],
head_type=cfg['head_type'],
training=False,
cfg=cfg)
ckpt_path = tf.train.latest_checkpoint('./checkpoints/' + cfg['sub_name'])
if ckpt_path is not None:
print("[*] load ckpt from {}".format(ckpt_path))
model.load_weights(ckpt_path)
else:
print("[*] Cannot find ckpt from {}.".format(ckpt_path))
# exit()
if FLAGS.img_path:
print("[*] Encode {} to ./output_embeds.npy".format(FLAGS.img_path))
img = cv2.imread(FLAGS.img_path)
img = cv2.resize(img, (cfg['input_size'], cfg['input_size']))
img = img.astype(np.float32) / 255.
if len(img.shape) == 3:
img = np.expand_dims(img, 0)
embeds = l2_norm(model(img))
np.save('./output_embeds.npy', embeds)
else:
print("[*] Perform Retrieval Evaluation on Y.T.F and F.S...")
# mAp_ytf, rr_ytf = perform_val_yts(cfg['eval_batch_size'], model, cfg['test_dataset_ytf'],img_ext='jpg')
# mAp_fs, rr_fs = perform_val_yts(cfg['eval_batch_size'], model, cfg['test_dataset_fs'],img_ext='png')
# print(" Y.T.F mAP {:.4f}, F.S mAP: {:.2f}".format(mAp_ytf, mAp_fs))
# print(" Y.T.F CMC-1 {:.4f}, F.S CMC-1: {:.2f}".format(rr_ytf[0], rr_fs[0]))
mAp_fs = mAp_ytf = 0
rr_ytf = rr_fs = [0]
is_flip = False
print('[*] is_flip : {}'.format(is_flip))
print("[*] Loading LFW, AgeDB30 and CFP-FP...")
lfw, agedb_30, cfp_fp, lfw_issame, agedb_30_issame, cfp_fp_issame = \
get_val_data(cfg['test_dataset'])
print("[*] Perform Evaluation on LFW...")
acc_lfw, best_th_lfw, auc_lfw, eer_lfw, embeddings_lfw = perform_val(
cfg['embd_shape'], cfg['batch_size'], model, lfw, lfw_issame,
is_ccrop=cfg['is_ccrop'], cfg=cfg,measure='Cosine',is_flip=is_flip)
print(" acc {:.4f}, th: {:.2f}, auc {:.4f}, EER {:.4f}".format(acc_lfw, best_th_lfw, auc_lfw, eer_lfw))
with open('embeddings/' + cfg['sub_name'] + '_embeddings_orig_lfw.csv', 'w', newline='') as file:
writer = csv.writer(file, escapechar='/', quoting=csv.QUOTE_NONE)
writer.writerows(embeddings_lfw)
print("[*] Perform Evaluation on AgeDB30...")
acc_agedb30, best_th_agedb30, auc_agedb30, eer_agedb30, embeddings_agedb30 = perform_val(
cfg['embd_shape'], cfg['batch_size'], model, agedb_30,
agedb_30_issame, is_ccrop=cfg['is_ccrop'], cfg=cfg,measure='Cosine',is_flip=is_flip)
print(" acc {:.4f}, th: {:.2f}, auc {:.4f}, EER {:.4f}".format(acc_agedb30, best_th_agedb30, auc_agedb30,
eer_agedb30))
print("[*] Perform Evaluation on CFP-FP...")
acc_cfp_fp, best_th_cfp_fp, auc_cfp_fp, eer_cfp_fp, embeddings_cfp_fp = perform_val(
cfg['embd_shape'], cfg['batch_size'], model, cfp_fp, cfp_fp_issame,
is_ccrop=cfg['is_ccrop'], cfg=cfg,measure='Cosine',is_flip=is_flip)
print(" acc {:.4f}, th: {:.2f}, auc {:.4f}, EER {:.4f}".format(acc_cfp_fp, best_th_cfp_fp, auc_cfp_fp,
eer_cfp_fp))
log_str = '''| q = {:.2f}, m = {:.2f} | LFW | AgeDB30 | CFP - FP |
|------------------------|--------|---------|----------|
| Accuracy | {:.4f} | {:.4f} | {:.4f} |
| EER | {:.4f} | {:.4f} | {:.4f} |
| AUC | {:.4f} | {:.4f} | {:.4f} |
| Threshold | {:.4f} | {:.4f} | {:.4f} |
| | mAP | CMC-1 | |
| Y.T.F | {:.4f} | {:.4f} | |
| F.S | {:.4f} | {:.4f} | | '''.format(0, 0,
acc_lfw, acc_agedb30, acc_cfp_fp,
eer_lfw, eer_agedb30, eer_cfp_fp,
auc_lfw, auc_agedb30, auc_cfp_fp,
best_th_lfw, best_th_agedb30,
best_th_cfp_fp,
mAp_ytf, rr_ytf[0],
mAp_fs, rr_fs[0])
with open('./logs/' + cfg['sub_name'] + "Output.txt", "a") as text_file:
text_file.write(log_str)
print(log_str)
log_str2 = '''| q = {:.2f}, m = {:.2f},LUT={},dist={} \t {:.4f} \t {:.4f} \t {:.4f} \t {:.4f} \t {:.4f} \t {:.4f} \t {:.4f} \t {:.4f} \t {:.4f} \t {:.4f} \t {:.4f} \t {:.4f} \t {:.4f}\n\n '''.format(
0, 0, 0, 0, mAp_ytf, mAp_fs, rr_ytf[0], rr_fs[0], eer_lfw, eer_agedb30, eer_cfp_fp, acc_lfw,
acc_agedb30, acc_cfp_fp, auc_lfw, auc_agedb30, auc_cfp_fp)
print(log_str2)
with open('./logs/' + cfg['sub_name'] + "_Output.md", "a") as text_file:
text_file.write(log_str2)
if __name__ == '__main__':
try:
app.run(main)
except SystemExit:
pass