forked from nianticlabs/stereo-from-mono
-
Notifications
You must be signed in to change notification settings - Fork 0
/
options.py
executable file
·144 lines (116 loc) · 6.08 KB
/
options.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
import argparse
class Options:
def __init__(self):
self.options = None
self.parser = argparse.ArgumentParser()
# Universal Options
self.parser.add_argument('--config_path',
help='path to config file containing dataset info',
type=str,
default='paths_config.yaml')
self.parser.add_argument('--mode',
help='training or inference mode',
type=str,
choices=['train', 'inference'],
default='train')
self.parser.add_argument('--height',
help='height of input images',
type=int,
default=320)
self.parser.add_argument('--width',
help='width of input images',
type=int,
default=608)
self.parser.add_argument('--disable_synthetic_augmentation',
action='store_true')
# Network Options
self.parser.add_argument('--network',
choices=['hourglass'],
default='hourglass')
self.parser.add_argument('--max_disparity',
help='maximum disparity',
type=int,
default=192)
self.parser.add_argument('--psm_no_SPP',
help='whether to use spatial pyramid pooling from PSM',
action='store_true')
self.parser.add_argument('--big_SPP',
help='standard PSM SPP module breaks for lower resolution images,'
'so by default we use smaller windows',
action='store_true')
self.parser.add_argument('--disable_normalisation',
action='store_true')
self.parser.add_argument('--disable_sharpening',
action='store_true')
self.parser.add_argument('--disable_background',
action='store_true')
self.parser.add_argument('--monodepth_model',
type=str,
default='midas',
choices=['midas', 'megadepth'])
self.parser.add_argument('--data_sampling',
type=float,
default=1.0)
# Training Options
self.parser.add_argument('--training_datasets',
help='datasets to train from',
nargs='+',
choices=['ADE20K', 'diode', 'diw', 'mapillary', 'mscoco',
'sceneflow', 'kitti2015'],
default=['ADE20K', 'diode', 'diw', 'mapillary', 'mscoco'])
self.parser.add_argument('--training_steps',
help='number of steps to train for',
type=int,
default=250000)
self.parser.add_argument('--log_freq',
help='sets the frequency of logs to tensorboard',
type=int,
default=250)
self.parser.add_argument('--val_batches',
help='number of validation batches to run and average over',
type=int,
default=1)
self.parser.add_argument('--batch_size',
help='number of images in each batch',
type=int,
default=2)
self.parser.add_argument('--lr',
help='the learning rate',
type=float,
default=1e-3)
self.parser.add_argument('--lr_step_size',
type=int,
default=5)
self.parser.add_argument('--num_workers',
help=' number of workers for dataloaders',
type=int,
default=6)
self.parser.add_argument('--model_name',
help='the name of the model for saving',
type=str,
default='model')
self.parser.add_argument('--log_path',
help='the path to save tensorboard events and trained models to',
type=str,
default='./logs')
self.parser.add_argument('--start_step',
help='step in training to start from - allows continuing from'
'loaded model',
default=0,
type=int)
# Test Options
self.parser.add_argument('--load_path',
help='the model path to load from',
type=str)
self.parser.add_argument('--save_disparities',
help='if set, save all computed disparities',
action='store_true')
self.parser.add_argument('--test_data_types',
choices=['eth3d', 'middlebury', 'kitti2015', 'kitti2012',
'flicker', 'kitti2015submission', 'sceneflow'],
nargs='+',
default=['eth3d', 'middlebury', 'kitti2015', 'kitti2012'])
def parse(self):
""" Parse arguments """
self.options = self.parser.parse_args()
return self.options