LAN/evaluate_model.py at master · draimundo/LAN · GitHub

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
###########################################################
# Evaluate the trained model on numerical testing patches #
###########################################################


import tensorflow as tf
import numpy as np
from datetime import datetime
from tqdm import tqdm

from load_dataset import load_test_data
from model import lan_g
import utils
import vgg
import sys

import lpips_tf

dataset_dir, vgg_dir, dslr_dir, phone_dir, model_dir,\
    restore_iter, img_h, img_w, batch_size, use_gpu = utils.process_evaluate_model_args(sys.argv)


FAC_PATCH = 1
PATCH_DEPTH = 1

PATCH_WIDTH = 256//FAC_PATCH
PATCH_HEIGHT = 256//FAC_PATCH
TARGET_WIDTH = int(PATCH_WIDTH * FAC_PATCH)
TARGET_HEIGHT = int(PATCH_HEIGHT * FAC_PATCH)
TARGET_DEPTH = 3
TARGET_SIZE = TARGET_WIDTH * TARGET_HEIGHT * TARGET_DEPTH

vgg_dir = 'vgg_pretrained/imagenet-vgg-verydeep-19.mat'

print("Loading testing data...")
test_data, test_answ = load_test_data(dataset_dir, dslr_dir, phone_dir, PATCH_WIDTH, PATCH_HEIGHT)
print("Testing data was loaded\n")

TEST_SIZE = test_data.shape[0]
num_test_batches = int(test_data.shape[0] / batch_size)

time_start = datetime.now()

config = tf.compat.v1.ConfigProto(device_count={'GPU': 0}) if not use_gpu else None
with tf.compat.v1.Session(config=config) as sess:
    phone_ = tf.compat.v1.placeholder(tf.float32, [batch_size, PATCH_HEIGHT, PATCH_WIDTH, PATCH_DEPTH])
    dslr_ = tf.compat.v1.placeholder(tf.float32, [batch_size, TARGET_HEIGHT, TARGET_WIDTH, TARGET_DEPTH])

    # Get the processed enhanced image
    enhanced = lan_g(phone_)

    saver = tf.compat.v1.train.Saver()

    dslr_gray = tf.image.rgb_to_grayscale(dslr_)
    enhanced_gray = tf.image.rgb_to_grayscale(enhanced)

    ## PSNR loss
    loss_psnr = tf.reduce_mean(tf.image.psnr(enhanced, dslr_, 1.0))
    loss_list = [loss_psnr]
    loss_text = ["loss_psnr"]

    # MSE loss
    loss_mse = tf.reduce_mean(tf.math.squared_difference(enhanced, dslr_))
    loss_list.append(loss_mse)
    loss_text.append("loss_mse")

    ## L1 loss
    loss_l1 = tf.reduce_mean(tf.abs(tf.math.subtract(enhanced, dslr_)))
    loss_list.append(loss_l1)
    loss_text.append("loss_l1")

    ## SSIM loss
    loss_ssim = 1 - tf.reduce_mean(tf.image.ssim(enhanced_gray, dslr_gray, 1.0))
    loss_list.append(loss_ssim)
    loss_text.append("loss_ssim")

    # MS-SSIM loss
    loss_ms_ssim = 1 - tf.reduce_mean(tf.image.ssim_multiscale(enhanced_gray, dslr_gray, 1.0))
    loss_list.append(loss_ms_ssim)
    loss_text.append("loss_ms_ssim")

    ## Content loss
    CONTENT_LAYER = 'relu5_4'
    enhanced_vgg = vgg.net(vgg_dir, vgg.preprocess(enhanced * 255))
    dslr_vgg = vgg.net(vgg_dir, vgg.preprocess(dslr_ * 255))
    loss_content = tf.reduce_mean(tf.math.squared_difference(enhanced_vgg[CONTENT_LAYER], dslr_vgg[CONTENT_LAYER]))
    loss_list.append(loss_content)
    loss_text.append("loss_content")

    ## LPIPS
    loss_lpips = tf.reduce_mean(lpips_tf.lpips(enhanced, dslr_, net='vgg'))
    loss_list.append(loss_lpips)
    loss_text.append("loss_lpips")

    ## Huber loss
    delta = 1
    abs_error = tf.abs(tf.math.subtract(enhanced, dslr_))
    quadratic = tf.math.minimum(abs_error, delta)
    linear = tf.math.subtract(abs_error, quadratic)
    loss_huber = tf.reduce_mean(0.5*tf.math.square(quadratic)+linear)
    loss_list.append(loss_huber)
    loss_text.append("loss_huber")

    ## Total variation loss
    loss_tv = tf.reduce_mean(tf.image.total_variation(enhanced))
    loss_list.append(loss_tv)
    loss_text.append("loss_tv")

    saver.restore(sess, model_dir + "LAN_iteration_" + str(restore_iter) + ".ckpt")
    test_losses_gen = np.zeros((1, len(loss_text)))
    for j in tqdm(range(num_test_batches)):

        be = j * batch_size
        en = (j+1) * batch_size

        phone_images = test_data[be:en]
        dslr_images = test_answ[be:en]

        [losses, enhanced_images] = sess.run([loss_list, enhanced], feed_dict={phone_: phone_images, dslr_: dslr_images})
        test_losses_gen += np.asarray(losses) / num_test_batches

    logs_gen = "Losses - iter: " + str(restore_iter) + "-> "
    for idx, loss in enumerate(loss_text):
        logs_gen += "%s: %.6g; " % (loss, test_losses_gen[0][idx])
    logs_gen += '\n'
    print(logs_gen)

    logs = open(model_dir + model_dir.split('/')[0] + ".txt", "a+")
    logs.write(logs_gen)
    logs.write('\n')
    logs.close()
print('total test time:', datetime.now() - time_start)