-
Notifications
You must be signed in to change notification settings - Fork 1
/
run_sample.py
86 lines (68 loc) · 3.45 KB
/
run_sample.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
from core.libs import pyutils
import argparse
import torch
import numpy as np
import random
import os
def fix_seed(seed=0):
torch.manual_seed(seed)
torch.cuda.manual_seed_all(seed)
random.seed(seed)
np.random.seed(seed)
# torch.set_deterministic(True)
if __name__ == '__main__':
parser = argparse.ArgumentParser()
parser.add_argument("--seed", default=5, type=int)
# Class Activation Map
parser.add_argument("--cam_network", default="net.resnet50_cam", type=str)
# parser.add_argument("--cam_crop_size", default=512, type=int)
# parser.add_argument("--cam_batch_size", default=16, type=int)
# parser.add_argument("--cam_crop_size", default=352, type=int)
parser.add_argument("--cam_batch_size", default=16, type=int)
parser.add_argument("--cam_num_epoches", default=2, type=int)
parser.add_argument("--cam_learning_rate", default=0.1, type=float)
parser.add_argument("--cam_weight_decay", default=1e-4, type=float)
parser.add_argument("--cam_eval_thres", default=0.11, type=float)
parser.add_argument("--cam_scales", default=(1.0,),
help="Multi-scale inferences")
parser.add_argument("--conf_fg_thres", default=0.20, type=float) # default=0.30
parser.add_argument("--conf_bg_thres", default=0.07, type=float)
# parameters for 'make_boundary_label'
parser.add_argument("--window_size", default=13, type=int)
parser.add_argument("--theta_scale", default=0.30, type=float)
parser.add_argument("--theta_diff", default=0.10, type=float)
parser.add_argument("--boundary_labels",
default={'BG': 0, 'FG': 50, 'BOUNDARY_FG_FG': 100, 'BOUNDARY_FG_BG': 150, 'IGNORE': 200},
type=dict)
# Output Path
parser.add_argument("--log_name", default="record", type=str)
parser.add_argument("--cam_weights_name", default="outputs/weights/sess/res50_cam_attention.pth", type=str)
parser.add_argument("--bes_weights_name", default="outputs/weights/sess/res50_bes.pth", type=str)
parser.add_argument("--cam_out_dir", default="test_results/cam", type=str)
parser.add_argument("--cam_vis_dir", default="test_results/cam_vis", type=str)
parser.add_argument("--boundary_label_dir", default="test_results/boundary_label", type=str)
parser.add_argument("--sem_seg_out_dir", default="test_results/sem_seg", type=str)
# Step
parser.add_argument("--train_cam_pass", default=False)
parser.add_argument("--make_cam_pass", default=False)
parser.add_argument("--make_boundary_label_pass", default=True)
args = parser.parse_args()
fix_seed(args.seed)
os.makedirs("outputs/weight/sess", exist_ok=True)
os.makedirs(args.cam_out_dir, exist_ok=True)
os.makedirs(args.cam_vis_dir, exist_ok=True)
os.makedirs(args.boundary_label_dir, exist_ok=True)
os.makedirs(args.sem_seg_out_dir, exist_ok=True)
print(vars(args))
if args.train_cam_pass is True:
from core.train import train_cam
timer = pyutils.Timer('step.train_cam:')
train_cam.run(args)
if args.make_cam_pass is True:
from core.train import make_cam
timer = pyutils.Timer('step.make_cam:')
make_cam.run(args)
if args.make_boundary_label_pass is True:
from core.train import make_boundary_label
timer = pyutils.Timer('step.cam_to_boundary_label:')
make_boundary_label.run(args)