-
Notifications
You must be signed in to change notification settings - Fork 0
/
rest_service.py
139 lines (91 loc) · 5.16 KB
/
rest_service.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
from __future__ import print_function
import tensorflow as tf
import keras.backend as K
import numpy as np
from keras.models import Sequential
from keras.layers import Conv2D, Deconv2D
import scipy
import matplotlib.pyplot as plt
import cv2
import os
print(os.getcwd())
project_dir = os.getcwd()
print(project_dir)
sess = tf.Session()
K.set_session(sess)
from flask import Flask, request
from flask_cors import CORS
from flask import send_file
app = Flask(__name__)
CORS(app)
def completion_function_eval(original, mask):
generated_image = generation(original)
return generated_image
with tf.name_scope('Generator_Model'):
generation = Sequential(name='Generation_Model')
generation.add(Conv2D(filters=64, kernel_size=(5,5), activation='relu',strides=(1,1), input_shape=(256,256,3), padding='same', name='Conv_1'))
generation.add(Conv2D(filters=128, kernel_size=(3,3), activation='relu', strides=(2,2), name='Conv_2', padding='same'))
generation.add(Conv2D(filters=128, kernel_size=(3,3), activation='relu', strides=(1,1), name='Conv_3', padding='same'))
generation.add(Conv2D(filters=256, kernel_size=(3,3), activation='relu', strides=(2,2), name='Conv_4', padding='same'))
generation.add(Conv2D(filters=256, kernel_size=(3,3), activation='relu', strides=(1,1), name='Conv_5', padding='same'))
generation.add(Conv2D(filters=256, kernel_size=(3,3), activation='relu', strides=(1,1), name='Conv_6', padding='same'))
generation.add(Conv2D(filters=256, kernel_size=(3,3), activation='relu', strides=(1,1), name='Dilated_Conv_1', dilation_rate=2, padding='same'))
generation.add(Conv2D(filters=256, kernel_size=(3,3), activation='relu', strides=(1,1), name='Dilated_Conv_2', dilation_rate=4, padding='same'))
generation.add(Conv2D(filters=256, kernel_size=(3,3), activation='relu', strides=(1,1), name='Dilated_Conv_3', dilation_rate=8, padding='same'))
generation.add(Conv2D(filters=256, kernel_size=(3,3), activation='relu', strides=(1,1), name='Dilated_Conv_4', dilation_rate=16, padding='same'))
generation.add(Conv2D(filters=256, kernel_size=(3,3), activation='relu', strides=(1,1), name='Conv_7', padding='same'))
generation.add(Conv2D(filters=256, kernel_size=(3,3), activation='relu', strides=(1,1), name='Conv_8', padding='same'))
generation.add(Deconv2D(filters=128, kernel_size=(4,4), activation='relu', strides=(2,2), name='DeConv_1', padding='same'))
generation.add(Conv2D(filters=128, kernel_size=(3,3), activation='relu', strides=(1,1), name='Conv_9', padding='same'))
generation.add(Deconv2D(filters=64, kernel_size=(4,4), activation='relu', strides=(2,2), name='DeConv_2', padding='same'))
generation.add(Conv2D(filters=32, kernel_size=(3,3), activation='relu', strides=(1,1), name='Conv_10', padding='same'))
generation.add(Conv2D(filters=3, kernel_size=(3,3), strides=(1,1), padding='same', name='Output'))
with tf.name_scope('ph_patched_batch'):
patched_batch_placeholder = K.placeholder(shape=(None, 256, 256, 3), dtype=tf.float32, name='ph_patched_batch')
with tf.name_scope('ph_mask_'):
mask_placeholder = K.placeholder(shape=(None, 256, 256, 3), dtype=tf.float32, name='ph_mask')
with tf.name_scope('Generator_Network'):
generated_batch = generation(patched_batch_placeholder)
with tf.name_scope('Generated_batchX1000'):
generated_batch = generated_batch*1
loss_dir_name = 'main_session'
init_op = tf.global_variables_initializer()
sess.run(init_op)
saver = tf.train.Saver(keep_checkpoint_every_n_hours=0.1)
# # Restoring the main Session if it exists
if os.path.isfile('saved_sessions/' + loss_dir_name + '/session.index'):
saver.restore(sess, 'saved_sessions/' + loss_dir_name + '/session')
else:
print('No session found!!')
# Testing with van data
import base64
from PIL import Image
import io
def stringToImage(base64_string):
imgdata = base64.b64decode(base64_string)
return Image.open(io.BytesIO(imgdata))
@app.route('/scan/api', methods=['POST'])
def get_inpainted_image():
image = plt.imread(request.files['image'])
mask = plt.imread(request.files['mask'])
# mask = plt.imread(project_dir+'/data/custom_data/original/mask.png')
mask = mask[:, :, 0:3]
image = image[:, :, 0:3]
# image = plt.imread(os.getcwd()+'/data/custom_data/original/image.png')
image = cv2.resize(image, (256, 256))
mask = cv2.resize(mask, (256, 256))
image= np.expand_dims(image,0)
mask = np.expand_dims(mask, 0)
mean = 0
mean_image = np.ones(np.shape(image))*mean
feed_image = mean_image*mask
feed_image = (1-mask)*image + (mask)*feed_image
test_custom = completion_function_eval(patched_batch_placeholder, mask_placeholder)
with sess.as_default():
test_custom_value = test_custom.eval(feed_dict={patched_batch_placeholder: feed_image,
mask_placeholder: mask})
image = (1-mask)*feed_image + mask*test_custom_value[0]
scipy.misc.imsave(project_dir+'/data/custom_data/image_result.png', image[0, :, :, :])
return send_file(project_dir+'/data/custom_data/image_result.png', mimetype='image/gif')
if __name__ == '__main__':
app.run(debug=True, host='0.0.0.0', port=8000)