-
Notifications
You must be signed in to change notification settings - Fork 9
/
train_thu.py
184 lines (149 loc) · 9.33 KB
/
train_thu.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
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
from tqdm import tqdm
import numpy as np
import torch
def train(args, model, dataloader, pair_dataloader, criterion, optimizer):
model.train()
print("-------------------------------------------------------------------------------")
device = args.device
# train_process
train_num_correct = 0
train_num_total = 0
loss_stack = []
acm_loss_stack = []
act_inst_loss_stack = []
act_cont_loss_stack = []
act_back_loss_stack = []
guide_loss_stack = []
att_loss_stack = []
feat_loss_stack = []
lcs_loss_stack = []
fsd_loss_stack = []
if not args.ftcl:
for input_feature, vid_label_t in tqdm(dataloader):
vid_label_t = vid_label_t.to(device)
input_feature = input_feature.to(device)
act_inst_cls, act_cont_cls, act_back_cls, \
act_inst_feat, act_cont_feat, act_back_feat, \
temp_att, act_inst_cas, _, _, _, \
lcs_candi, fsd_act_candi, fsd_bak_candi = model(input_feature)
loss, loss_dict = criterion(act_inst_cls, act_cont_cls, act_back_cls, vid_label_t, temp_att,
act_inst_feat, act_cont_feat, act_back_feat, act_inst_cas,
lcs_candi, fsd_act_candi, fsd_bak_candi, args)
optimizer.zero_grad()
if not torch.isnan(loss):
loss.backward()
optimizer.step()
with torch.no_grad():
fg_score = act_inst_cls[:, :args.action_cls_num]
label_np = vid_label_t.cpu().numpy()
score_np = fg_score.cpu().numpy()
pred_np = np.zeros_like(score_np)
pred_np[score_np >= args.cls_threshold] = 1
pred_np[score_np < args.cls_threshold] = 0
correct_pred = np.sum(label_np == pred_np, axis=1)
train_num_correct += np.sum((correct_pred == args.action_cls_num))
train_num_total += correct_pred.shape[0]
loss_stack.append(loss.cpu().item())
act_inst_loss_stack.append(loss_dict["act_inst_loss"])
act_cont_loss_stack.append(loss_dict["act_cont_loss"])
act_back_loss_stack.append(loss_dict["act_back_loss"])
guide_loss_stack.append(loss_dict["guide_loss"])
feat_loss_stack.append(loss_dict["feat_loss"])
att_loss_stack.append(loss_dict["sparse_loss"])
acm_loss_stack.append(loss_dict["acm_loss"])
lcs_loss_stack.append(loss_dict["lcs_loss"])
fsd_loss_stack.append(loss_dict["fsd_loss"])
train_acc = train_num_correct / train_num_total
train_log_dict = {"train_act_inst_cls_loss": np.mean(act_inst_loss_stack),
"train_act_cont_cls_loss": np.mean(act_cont_loss_stack),
"train_act_back_cls_loss": np.mean(act_back_loss_stack),
"train_guide_loss": np.mean(guide_loss_stack),
"train_feat_loss": np.mean(feat_loss_stack),
"train_att_loss": np.mean(att_loss_stack),
"train_acm_loss": np.mean(acm_loss_stack),
"train_lcs_loss": np.mean(lcs_loss_stack),
"train_fsd_loss": np.mean(fsd_loss_stack),
"train_loss": np.mean(loss_stack),
"train_acc": train_acc}
print("")
print("train_act_inst_cls_loss:{:.3f} train_act_cont_cls_loss:{:.3f}".format(np.mean(act_inst_loss_stack),
np.mean(act_cont_loss_stack)))
print("train_act_back_cls_loss:{:.3f} train_att_loss:{:.3f}".format(np.mean(act_back_loss_stack),
np.mean(att_loss_stack)))
print("train_feat_loss: {:.3f} train_loss:{:.3f}".format(np.mean(feat_loss_stack), np.mean(loss_stack)))
print("train acc:{:.3f}".format(train_acc))
print("-------------------------------------------------------------------------------")
return train_log_dict
else:
for input_feature_1, input_feature_2, vid_label_1, vid_label_2 in tqdm(pair_dataloader):
vid_label_1 = vid_label_1.to(device)
vid_label_2 = vid_label_2.to(device)
input_feature_1 = input_feature_1.to(device)
input_feature_2 = input_feature_2.to(device)
output_1, output_2 = model(args.ftcl, input_feature_1, input_feature_2)
act_inst_cls_1, act_cont_cls_1, act_back_cls_1, act_inst_feat_1, act_cont_feat_1, act_back_feat_1, \
temp_att_1, act_inst_cas_1, act_cas_1, act_cont_cas_1, act_back_cas_1, \
candi_for_dp_1, act_candi_for_nw_1, bak_candi_for_nw_1 = output_1
act_inst_cls_2, act_cont_cls_2, act_back_cls_2, act_inst_feat_2, act_cont_feat_2, act_back_feat_2, \
temp_att_2, act_inst_cas_2, act_cas_2, act_cont_cas_2, act_back_cas_2, \
candi_for_dp_2, act_candi_for_nw_2, bak_candi_for_nw_2 = output_2
loss, loss_dict = criterion(act_inst_cls_1, act_cont_cls_1, act_back_cls_1, vid_label_1, temp_att_1,
act_inst_feat_1, act_cont_feat_1, act_back_feat_1, act_inst_cas_1,
candi_for_dp_1, act_candi_for_nw_1, bak_candi_for_nw_1,
args,
act_inst_cls_2, act_cont_cls_2, act_back_cls_2, vid_label_2, temp_att_2,
act_inst_feat_2, act_cont_feat_2, act_back_feat_2, act_inst_cas_2,
candi_for_dp_2, act_candi_for_nw_2, bak_candi_for_nw_2,
)
optimizer.zero_grad()
if not torch.isnan(loss):
loss.backward()
optimizer.step()
with torch.no_grad():
fg_score_1 = act_inst_cls_1[:, :args.action_cls_num]
fg_score_2 = act_inst_cls_2[:, :args.action_cls_num]
label_np_1 = vid_label_1.cpu().numpy()
label_np_2 = vid_label_2.cpu().numpy()
score_np_1 = fg_score_1.cpu().numpy()
score_np_2 = fg_score_2.cpu().numpy()
pred_np_1 = np.zeros_like(score_np_1)
pred_np_2 = np.zeros_like(score_np_2)
pred_np_1[score_np_1 >= args.cls_threshold] = 1
pred_np_2[score_np_2 >= args.cls_threshold] = 1
pred_np_1[score_np_1 < args.cls_threshold] = 0
pred_np_2[score_np_2 < args.cls_threshold] = 0
correct_pred_1 = np.sum(label_np_1 == pred_np_1, axis=1)
correct_pred_2 = np.sum(label_np_2 == pred_np_2, axis=1)
train_num_correct += np.sum(((correct_pred_1 == args.action_cls_num) *
(correct_pred_2 == args.action_cls_num)))
train_num_total += correct_pred_1.shape[0]
loss_stack.append(loss.cpu().item())
act_inst_loss_stack.append(loss_dict["act_inst_loss"])
act_cont_loss_stack.append(loss_dict["act_cont_loss"])
act_back_loss_stack.append(loss_dict["act_back_loss"])
guide_loss_stack.append(loss_dict["guide_loss"])
feat_loss_stack.append(loss_dict["feat_loss"])
att_loss_stack.append(loss_dict["sparse_loss"])
acm_loss_stack.append(loss_dict["acm_loss"])
lcs_loss_stack.append(loss_dict["lcs_loss"])
fsd_loss_stack.append(loss_dict["fsd_loss"])
train_acc = train_num_correct / train_num_total
train_log_dict = {"train_act_inst_cls_loss": np.mean(act_inst_loss_stack),
"train_act_cont_cls_loss": np.mean(act_cont_loss_stack),
"train_act_back_cls_loss": np.mean(act_back_loss_stack),
"train_guide_loss": np.mean(guide_loss_stack),
"train_feat_loss": np.mean(feat_loss_stack),
"train_att_loss": np.mean(att_loss_stack),
"train_acm_loss": np.mean(acm_loss_stack),
"train_lcs_loss": np.mean(lcs_loss_stack),
"train_fsd_loss": np.mean(fsd_loss_stack),
"train_loss": np.mean(loss_stack),
"train_acc": train_acc}
print("\n")
print("train_act_inst_cls_loss:{:.3f} train_act_cont_cls_loss:{:.3f}".format(np.mean(act_inst_loss_stack),
np.mean(act_cont_loss_stack)))
print("train_act_back_cls_loss:{:.3f} train_att_loss:{:.3f}".format(np.mean(act_back_loss_stack),
np.mean(att_loss_stack)))
print("train_feat_loss: {:.3f} train_loss:{:.3f}".format(np.mean(feat_loss_stack), np.mean(loss_stack)))
print("train acc:{:.3f}".format(train_acc))
return train_log_dict