-
Notifications
You must be signed in to change notification settings - Fork 0
/
Copy pathtest.py
193 lines (154 loc) · 8.39 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
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
import os.path
from sr_model.capsule_srgan import generator
from trainer.utils import load_image, plot_sample
from sr_model import resolve_single
import tensorflow as tf
import datetime
import numpy as np
import csv
subset = "RANDOM_BLUR"
def run_test(generator_model, output_name_suffix, lr, hr, output_dir):
# Load LR image
lr_image = load_image(lr)
# Start timer
starting_time = datetime.datetime.now()
# Generate SR image
sr = resolve_single(generator_model, lr_image)
# End timer
end = datetime.datetime.now()
# Calculate time
time_taken = str((end - starting_time).total_seconds())
# Calculate PSNR
ground_truth = load_image(hr)
expand_ground_truth = ground_truth[:, :, np.newaxis]
psnr_value = tf.image.psnr(tf.expand_dims(expand_ground_truth, axis=0), sr, max_val=255)[0]
# Save image to destination
src_png = tf.image.encode_png(sr)
split_filename = os.path.splitext(os.path.basename(lr))
output_name = split_filename[0] + output_name_suffix + split_filename[1]
output_path = os.path.join(output_dir, output_name)
tf.io.write_file(tf.constant(output_path), src_png)
return time_taken, psnr_value
allfieldnames = ['file_name', 'CapsSRGAN_NoBlur_PSNR', 'CapsSRGAN_NoBlur_RunTime', 'CapsSRGAN_RandomBlur_PSNR',
'CapsSRGAN_RandomBlur_RunTime', 'OriginalSRGAN_NoBlur_PSNR', 'OriginalSRGAN_NoBlur_RunTime',
'OriginalSRGAN_RandomBlur_PSNR', 'OriginalSRGAN_RandomBlur_RunTime', 'CapsSRGAN_AllBlur_PSNR',
'CapsSRGAN_AllBlur_RunTime', 'OriginalSRGAN_AllBlur_PSNR', 'OriginalSRGAN_AllBlur_RunTime']
fieldnames = ['file_name', 'CapsSRGAN_PSNR', 'CapsSRGAN_RunTime', 'OriginalSRGAN_PSNR', 'OriginalSRGAN_RunTime']
rows = []
caps_no_blur_generator_model = generator()
caps_no_blur_generator_model.load_weights('saved_model_weights/capsule_srgan/caps_gan_no_blur/caps_gan_generator.h5')
caps_random_blur_generator_model = generator()
caps_random_blur_generator_model.load_weights(
'saved_model_weights/Final/CapSRGAN/All_Blur/Generator.h5')
caps_all_blur_generator_model = generator()
caps_all_blur_generator_model.load_weights(
'saved_model_weights/capsule_srgan/caps_gan_random_blur/caps_gan_generator.h5')
original_no_blur_generator_model = generator()
original_no_blur_generator_model.load_weights(
'saved_model_weights/Original_srgan/Original_gan_no_blur/Original_gan_generator.h5')
original_random_blur_generator_model = generator()
original_random_blur_generator_model.load_weights(
'saved_model_weights/srgan/original_gan_random_blur/gan_generator_original.h5')
#
original_all_blur_generator_model = generator()
original_all_blur_generator_model.load_weights(
'saved_model_weights/Final/CapSRGAN/All_Blur/Generator.h5')
def test_all():
reference_dir = 'Dataset/Testing/Ground Truth/'
input_dir = 'Dataset/Testing/LR - With Blur/'
output_dir = 'Dataset/Testing/Results_With_Blur'
for file in os.listdir(input_dir):
input_image = os.path.join(input_dir, file)
reference_image = os.path.join(reference_dir, file)
# Test run
run_test(caps_no_blur_generator_model, "_CapsSRGAN_NoBlur", input_image, reference_image, output_dir)
''' Capsule SRGAN - No Blur Training'''
caps_srgan_no_blur_run_time, caps_srgan_no_blur_psnr = run_test(caps_no_blur_generator_model,
"_CapsSRGAN_NoBlur", input_image,
reference_image, output_dir)
caps_srgan_no_blur_psnr = f'{caps_srgan_no_blur_psnr.numpy():3f}'
print("Time taken for Capsule SRGAN - No Blur model : ", caps_srgan_no_blur_run_time)
print(f'PSNR = {caps_srgan_no_blur_psnr}')
''' Capsule SRGAN - Random Blur Training'''
caps_srgan_random_blur_run_time, caps_srgan_random_blur_psnr = run_test(caps_random_blur_generator_model,
"_CapsSRGAN_RandomBlur", input_image,
reference_image, output_dir)
caps_srgan_random_blur_psnr = f'{caps_srgan_random_blur_psnr.numpy():3f}'
print("Time taken for Capsule SRGAN - Random Blur model : ", caps_srgan_random_blur_run_time)
print(f'PSNR = {caps_srgan_random_blur_psnr}')
''' Original SRGAN - No Blur Training'''
original_srgan_no_blur_run_time, original_srgan_no_blur_psnr = run_test(original_no_blur_generator_model,
"_OriginalSRGAN_NoBlur", input_image,
reference_image, output_dir)
original_srgan_no_blur_psnr = f'{original_srgan_no_blur_psnr.numpy():3f}'
print("Time taken for Original SRGAN - No Blur model : ", original_srgan_no_blur_run_time)
print(f'PSNR = {original_srgan_no_blur_psnr}')
''' Original SRGAN - Random Blur Training'''
original_srgan_random_blur_run_time, original_srgan_random_blur_psnr = run_test(
original_random_blur_generator_model,
"_OriginalSRGAN_RandomBlur", input_image, reference_image, output_dir)
original_srgan_random_blur_psnr = f'{original_srgan_random_blur_psnr.numpy():3f}'
print("Time taken for Original SRGAN - Random Blur model : ", original_srgan_random_blur_run_time)
print(f'PSNR = {original_srgan_random_blur_psnr}')
rows.append({
'file_name': input_image,
'CapsSRGAN_NoBlur_PSNR': caps_srgan_no_blur_psnr,
'CapsSRGAN_NoBlur_RunTime': caps_srgan_no_blur_run_time,
'CapsSRGAN_RandomBlur_PSNR': caps_srgan_random_blur_psnr,
'CapsSRGAN_RandomBlur_RunTime': caps_srgan_random_blur_run_time,
'OriginalSRGAN_NoBlur_PSNR': original_srgan_no_blur_psnr,
'OriginalSRGAN_NoBlur_RunTime': original_srgan_no_blur_run_time,
'OriginalSRGAN_RandomBlur_PSNR': original_srgan_random_blur_psnr,
'OriginalSRGAN_RandomBlur_RunTime': original_srgan_random_blur_run_time
})
# plot_sample(lr, sr)
# Generate CSV Report
with open('test_results_with_blur.csv', 'w', encoding='UTF8', newline='') as f:
writer = csv.DictWriter(f, fieldnames=allfieldnames)
writer.writeheader()
writer.writerows(rows)
def test_model():
reference_dir = 'Dataset/Testing/Ground Truth/'
input_dir = 'Dataset/Testing/LR_' + subset + '/'
output_dir = 'Dataset/Testing/Final_Results/' + subset
for file in os.listdir(input_dir):
input_image = os.path.join(input_dir, file)
reference_image = os.path.join(reference_dir, file)
''' Capsule SRGAN '''
capsule_run_time, psnr = run_test(caps_random_blur_generator_model,
"", input_image,
reference_image,
(output_dir + "/Med-CapSRGAN_Results/"))
capsule_psnr = f'{psnr.numpy():3f}'
print("Time taken for Capsule SRGAN model : ", capsule_run_time)
print(f'PSNR = {capsule_psnr}')
''' Original SRGAN '''
original_run_time, psnr = run_test(
original_random_blur_generator_model,
"", input_image, reference_image, (output_dir + "/SRGAN_Results/"))
original_psnr = f'{psnr.numpy():3f}'
print("Time taken for Original SRGAN model : ", original_run_time)
print(f'PSNR = {original_psnr}')
rows.append({
'file_name': input_image,
'CapsSRGAN_PSNR': capsule_psnr,
'CapsSRGAN_RunTime': capsule_run_time,
'OriginalSRGAN_PSNR': original_psnr,
'OriginalSRGAN_RunTime': original_run_time
})
# Generate CSV Report
with open('test_results_' + subset + '.csv', 'w', encoding='UTF8', newline='') as f:
writer = csv.DictWriter(f, fieldnames=fieldnames)
writer.writeheader()
writer.writerows(rows)
def test_one(model):
input_image = 'Demo/1.png'
reference_image = 'Demo/Caps_Enhanced_1.png'
output_dir = 'Demo'
run_time, psnr = run_test(
model,
"_CapsEnhanced", input_image, reference_image, output_dir)
psnr = f'{psnr.numpy():3f}'
print("Time taken : ", run_time)
print(f'PSNR = {psnr}')
test_model()