-
Notifications
You must be signed in to change notification settings - Fork 3
/
inference_batch.py
160 lines (144 loc) · 7.13 KB
/
inference_batch.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
"""
Copyright (C) 2018 NVIDIA Corporation. All rights reserved.
Licensed under the CC BY-NC-SA 4.0 license (https://creativecommons.org/licenses/by-nc-sa/4.0/legalcode).
"""
from __future__ import print_function
from utils import get_config, get_data_loader_folder, pytorch03_to_pytorch04, load_inception
from trainer import MUNIT_Trainer, UNIT_Trainer
from torch import nn
from scipy.stats import entropy
import torch.nn.functional as F
import argparse
from torch.autograd import Variable
from data import ImageFolder
import numpy as np
import torchvision.utils as vutils
try:
from itertools import izip as zip
except ImportError: # will be 3.x series
pass
import sys
import torch
import os
parser = argparse.ArgumentParser()
parser.add_argument('--config', type=str, default='configs/edges2handbags_folder', help='Path to the config file.')
parser.add_argument('--input_folder', type=str, help="input image folder")
parser.add_argument('--output_folder', type=str, help="output image folder")
parser.add_argument('--checkpoint', type=str, help="checkpoint of autoencoders")
parser.add_argument('--a2b', type=int, help="1 for a2b and 0 for b2a", default=1)
parser.add_argument('--seed', type=int, default=1, help="random seed")
parser.add_argument('--num_style',type=int, default=30, help="number of styles to sample")
parser.add_argument('--synchronized', action='store_true', help="whether use synchronized style code or not")
parser.add_argument('--output_only', action='store_true', help="whether only save the output images or also save the input images")
parser.add_argument('--output_path', type=str, default='.', help="path for logs, checkpoints, and VGG model weight")
parser.add_argument('--trainer', type=str, default='MUNIT', help="MUNIT|UNIT")
parser.add_argument('--compute_IS', action='store_true', help="whether to compute Inception Score or not")
parser.add_argument('--compute_CIS', action='store_true', help="whether to compute Conditional Inception Score or not")
parser.add_argument('--inception_a', type=str, default='.', help="path to the pretrained inception network for domain A")
parser.add_argument('--inception_b', type=str, default='.', help="path to the pretrained inception network for domain B")
opts = parser.parse_args()
torch.manual_seed(opts.seed)
torch.cuda.manual_seed(opts.seed)
# Load experiment setting
config = get_config(opts.config)
input_dim = config['input_dim_a'] if opts.a2b else config['input_dim_b']
# Load the inception networks if we need to compute IS or CIIS
if opts.compute_IS or opts.compute_IS:
inception = load_inception(opts.inception_b) if opts.a2b else load_inception(opts.inception_a)
# freeze the inception models and set eval mode
inception.eval()
for param in inception.parameters():
param.requires_grad = False
inception_up = nn.Upsample(size=(299, 299), mode='bilinear')
# Setup model and data loader
image_names = ImageFolder(opts.input_folder, transform=None, return_paths=True)
data_loader = get_data_loader_folder(opts.input_folder, 1, False, new_size=None, crop=False, height=400, width=640)
config['vgg_model_path'] = opts.output_path
if opts.trainer == 'MUNIT':
style_dim = config['gen']['style_dim']
trainer = MUNIT_Trainer(config)
elif opts.trainer == 'UNIT':
trainer = UNIT_Trainer(config)
else:
sys.exit("Only support MUNIT|UNIT")
try:
state_dict = torch.load(opts.checkpoint)
trainer.gen_a.load_state_dict(state_dict['a'])
trainer.gen_b.load_state_dict(state_dict['b'])
except:
state_dict = pytorch03_to_pytorch04(torch.load(opts.checkpoint), opts.trainer)
trainer.gen_a.load_state_dict(state_dict['a'])
trainer.gen_b.load_state_dict(state_dict['b'])
trainer.cuda()
trainer.train()
encode = trainer.gen_a.encode if opts.a2b else trainer.gen_b.encode # encode function
decode = trainer.gen_b.decode if opts.a2b else trainer.gen_a.decode # decode function
if opts.compute_IS:
IS = []
all_preds = []
if opts.compute_CIS:
CIS = []
if opts.trainer == 'MUNIT':
# Start testing
style_fixed = Variable(torch.randn(opts.num_style, style_dim, 1, 1).cuda(), volatile=False)
for i, (images, names) in enumerate(zip(data_loader, image_names)):
if opts.compute_CIS:
cur_preds = []
print(names[1])
images = Variable(images.cuda(), volatile=True)
content, _ = encode(images)
style = style_fixed if opts.synchronized else Variable(torch.randn(opts.num_style, style_dim, 1, 1).cuda(), volatile=False)
for j in range(opts.num_style):
s = style[j].unsqueeze(0)
outputs = decode(content, s)
outputs = (outputs + 1) / 2.
if opts.compute_IS or opts.compute_CIS:
pred = F.softmax(inception(inception_up(outputs)), dim=1).cpu().data.numpy() # get the predicted class distribution
if opts.compute_IS:
all_preds.append(pred)
if opts.compute_CIS:
cur_preds.append(pred)
# path = os.path.join(opts.output_folder, 'input{:03d}_output{:03d}.jpg'.format(i, j))
basename = os.path.basename(names[1])
path = os.path.join(opts.output_folder+"_%02d"%j,basename)
if not os.path.exists(os.path.dirname(path)):
os.makedirs(os.path.dirname(path))
vutils.save_image(outputs.data, path, padding=0, normalize=True)
if opts.compute_CIS:
cur_preds = np.concatenate(cur_preds, 0)
py = np.sum(cur_preds, axis=0) # prior is computed from outputs given a specific input
for j in range(cur_preds.shape[0]):
pyx = cur_preds[j, :]
CIS.append(entropy(pyx, py))
if not opts.output_only:
# also save input images
vutils.save_image(images.data, os.path.join(opts.output_folder, 'input{:03d}.jpg'.format(i)), padding=0, normalize=True)
if opts.compute_IS:
all_preds = np.concatenate(all_preds, 0)
py = np.sum(all_preds, axis=0) # prior is computed from all outputs
for j in range(all_preds.shape[0]):
pyx = all_preds[j, :]
IS.append(entropy(pyx, py))
if opts.compute_IS:
print("Inception Score: {}".format(np.exp(np.mean(IS))))
if opts.compute_CIS:
print("conditional Inception Score: {}".format(np.exp(np.mean(CIS))))
elif opts.trainer == 'UNIT':
# Start testing
for i, (images, names) in enumerate(zip(data_loader, image_names)):
print(names[1])
images = Variable(images.cuda(), volatile=True)
content, _ = encode(images)
outputs = decode(content)
outputs = (outputs + 1) / 2.
# path = os.path.join(opts.output_folder, 'input{:03d}_output{:03d}.jpg'.format(i, j))
basename = os.path.basename(names[1])
path = os.path.join(opts.output_folder,basename)
if not os.path.exists(os.path.dirname(path)):
os.makedirs(os.path.dirname(path))
vutils.save_image(outputs.data, path, padding=0, normalize=True)
if not opts.output_only:
# also save input images
vutils.save_image(images.data, os.path.join(opts.output_folder, 'input{:03d}.jpg'.format(i)), padding=0, normalize=True)
else:
pass