-
Notifications
You must be signed in to change notification settings - Fork 0
/
train_baseline123_YCbCr_final.py
161 lines (112 loc) · 5.29 KB
/
train_baseline123_YCbCr_final.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
#!/usr/bin/python3
#coding=utf-8
import sys
sys.path.insert(0, '../')
import os
import torch
import torch.nn as nn
import torch.nn.functional as F
from torch.utils.data import DataLoader
from torch import optim
os.environ['CUDA_VISIBLE_DEVICES'] = '0'
from misc import AvgMeter, check_mkdir
from ORSI_SOD_dataset_YCbCr_final import ORSI_SOD_dataset
from src.baseline123_YCbCr import net as Net
from src.TFGM import AuxiliaryLoss
from evaluator_SR_YCbCr import Eval_thread
loss_enhance = AuxiliaryLoss()
args = {
'iter_num': 7500,
'epoch': 100,
'train_batch_size': 2,
'last_iter': 0,
'lr': 1e-3,
'lr_decay': 0.9,
'weight_decay': 0.0005,
'momentum': 0.9,
'snapshot': ''
}
torch.manual_seed(2021)
### saliency loss function
"""
smaps : BCE + wIOU
edges: BCE
"""
def structure_loss(pred, mask):
#mask = mask.detach()
wbce = F.binary_cross_entropy_with_logits(pred, mask)
pred = torch.sigmoid(pred)
inter = (pred*mask).sum(dim=(2,3))
union = (pred+mask).sum(dim=(2,3))
wiou = 1-(inter+1)/(union-inter+1)
return wbce.mean()+wiou.mean()#
#define dataset and dataloader
dataset = "ORS_4199" #or "ORSSD" or "EORSSD"
input_size = 224
train_set = ORSI_SOD_dataset(root = '/data/iopen/lyf/SaliencyOD_in_RSIs/'+ dataset +' dataset/', size = input_size, mode = "train", aug = True)
train_loader = DataLoader(train_set, batch_size = args['train_batch_size'], shuffle = True, num_workers = args['train_batch_size'])
test_set = ORSI_SOD_dataset(root = '/data/iopen/lyf/SaliencyOD_in_RSIs/'+ dataset +' dataset/', size = input_size, mode = "test", aug = False)
test_loader = DataLoader(test_set, batch_size = 1, shuffle = False, num_workers = 1)
args['iter_num'] = args["epoch"] * len(train_loader)
def main():
model = Net()
net = model.cuda().train()
optimizer = optim.SGD([
{'params': [param for name, param in net.named_parameters() if name[-4:] == 'bias'],
'lr': 2 * args['lr']},
{'params': [param for name, param in net.named_parameters() if name[-4:] != 'bias'],
'lr': args['lr'], 'weight_decay': args['weight_decay']}
], momentum=args['momentum'])
train(net, optimizer)
def train(net, optimizer):
curr_iter = args['last_iter']
for epoch in range(0, args['epoch']): # total 100 epoches
total_loss_record, t1_record, t2_record, t3_record = AvgMeter(), AvgMeter(), AvgMeter(), AvgMeter()
net.train()
for i, data in enumerate(train_loader):
optimizer.param_groups[0]['lr'] = 2 * args['lr'] * (1 - float(curr_iter) / args['iter_num'] #
) ** args['lr_decay']
optimizer.param_groups[1]['lr'] = 2 * args['lr'] * (1 - float(curr_iter) / args['iter_num']#
) ** args['lr_decay']
image_lr_rgb, img_YCbCr_lr, img_YCbCr_sr, label, name = data
label = label.cuda()
img_YCbCr_sr = img_YCbCr_sr.cuda()
input1 = image_lr_rgb.cuda()
input2 = img_YCbCr_lr.cuda()
optimizer.zero_grad()
smaps, pred_sr, sod_fea, sr_fea = net(input1, input2)
smap1, smap2, smap3, smap4, smap5 = smaps
########## compute loss #############
loss1_1 = structure_loss(smap1, label)
loss1_2 = structure_loss(smap2, label)
loss1_3 = structure_loss(smap3, label)
loss1_4 = structure_loss(smap4, label)
loss1_5 = structure_loss(smap5, label)
t1 = loss1_1 + loss1_2 + (loss1_3 / 2) + (loss1_4 / 4) + (loss1_5 / 8)
t2 = nn.MSELoss()(pred_sr, img_YCbCr_sr) ##input YCbCr，supervise YCbcr
t3 = loss_enhance(sod_fea*sr_fea, img_YCbCr_sr * torch.cat((label, label, label), dim=1))
if epoch == 0:
# warm up
total_loss = t1 + t2 + t3
else:
total_loss = t1 + 100*t2 + t3
total_loss.backward()
optimizer.step()
t1_record.update(t1.item(), args['train_batch_size'])
t2_record.update(t2.item(), args['train_batch_size'])
t3_record.update(t3.item(), args['train_batch_size'])
total_loss_record.update(total_loss.item(), args['train_batch_size'])
#############log###############
if curr_iter % 125 == 0:
log = '[epoch: %03d] [iter: %05d] [total loss %.5f] [loss1 %.8f] [loss2 %.8f] [loss3 %.8f] [lr %.13f] ' % \
(epoch, curr_iter, total_loss_record.avg, t1_record.avg, t2_record.avg, t3_record.avg, optimizer.param_groups[1]['lr'])
print(log)
curr_iter += 1
if epoch % 10 == 0 or (epoch >= 60 and epoch %2 ==0):
thread = Eval_thread(epoch = epoch, model = net.eval(), loader = test_loader, method = "baseline123_YCbCr_final_", dataset = dataset, output_dir = "./data/output", cuda=True)
logg, fm = thread.run()
print(logg)
torch.save(net.state_dict(), './data/model_224*224_bs=8_baseline123_YCbCr_final_'+ dataset +'/epoch_{}_{}.pth'.format(epoch,fm))
## #############end###############
if __name__ == '__main__':
main()