-
Notifications
You must be signed in to change notification settings - Fork 0
/
Copy pathjaad.py
166 lines (135 loc) · 8.05 KB
/
jaad.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
from utils.jaad_data import JAAD
from utils.jaad_preprocessing import *
import torch
from torch import nn
from torch.utils.data import TensorDataset, DataLoader
from torch.utils.tensorboard import SummaryWriter
from model.main_model import Model
from sklearn.metrics import accuracy_score, f1_score, roc_auc_score, precision_score, recall_score, confusion_matrix
import argparse
device = torch.device('cuda' if torch.cuda.is_available() else 'cpu')
def main(args):
if not args.learn:
seed_all(args.seed)
data_opts = {'fstride': 1,
'sample_type': args.bh, # 'beh'
'subset': 'default',
'height_rng': [0, float('inf')],
'squarify_ratio': 0,
'data_split_type': 'default', # kfold, random, default
'seq_type': 'crossing',
'min_track_size': 15,
'random_params': {'ratios': None,
'val_data': True,
'regen_data': False},
'kfold_params': {'num_folds': 5, 'fold': 1},
}
tte = [30, 60]
imdb = JAAD(data_path=args.set_path)
seq_train = imdb.generate_data_trajectory_sequence('train', **data_opts)
balanced_seq_train = balance_dataset(seq_train)
tte_seq_train, traj_seq_train = tte_dataset(balanced_seq_train, tte, 0.6, args.times_num)
seq_valid = imdb.generate_data_trajectory_sequence('val', **data_opts)
balanced_seq_valid = balance_dataset(seq_valid)
tte_seq_valid, traj_seq_valid = tte_dataset(balanced_seq_valid, tte, 0, args.times_num)
seq_test = imdb.generate_data_trajectory_sequence('test', **data_opts)
tte_seq_test, traj_seq_test = tte_dataset(seq_test, tte, 0, args.times_num)
bbox_train = tte_seq_train['bbox']
bbox_valid = tte_seq_valid['bbox']
bbox_test = tte_seq_test['bbox']
bbox_dec_train = traj_seq_train['bbox']
bbox_dec_valid = traj_seq_valid['bbox']
bbox_dec_test = traj_seq_test['bbox']
vel_train = tte_seq_train['vehicle_act']
vel_valid = tte_seq_valid['vehicle_act']
vel_test = tte_seq_test['vehicle_act']
action_train = tte_seq_train['activities']
action_valid = tte_seq_valid['activities']
action_test = tte_seq_test['activities']
normalized_bbox_train = normalize_bbox(bbox_train)
normalized_bbox_valid = normalize_bbox(bbox_valid)
normalized_bbox_test = normalize_bbox(bbox_test)
normalized_bbox_dec_train = normalize_traj(bbox_dec_train)
normalized_bbox_dec_valid = normalize_traj(bbox_dec_valid)
normalized_bbox_dec_test = normalize_traj(bbox_dec_test)
label_action_train = prepare_label(action_train)
label_action_valid = prepare_label(action_valid)
label_action_test = prepare_label(action_test)
X_train, X_valid = torch.Tensor(normalized_bbox_train), torch.Tensor(normalized_bbox_valid)
Y_train, Y_valid = torch.Tensor(label_action_train), torch.Tensor(label_action_valid)
X_test = torch.Tensor(normalized_bbox_test)
Y_test = torch.Tensor(label_action_test)
X_train_dec = torch.Tensor(pad_sequence(normalized_bbox_dec_train, 60))
X_valid_dec = torch.Tensor(pad_sequence(normalized_bbox_dec_valid, 60))
X_test_dec = torch.Tensor(pad_sequence(normalized_bbox_dec_test, 60))
vel_train = torch.Tensor(vel_train)
vel_valid = torch.Tensor(vel_valid)
vel_test = torch.Tensor(vel_test)
trainset = TensorDataset(X_train, Y_train, vel_train, X_train_dec)
validset = TensorDataset(X_valid, Y_valid, vel_valid, X_valid_dec)
testset = TensorDataset(X_test, Y_test, vel_test, X_test_dec)
train_loader = DataLoader(trainset, batch_size=args.batch_size, shuffle=True)
valid_loader = DataLoader(validset, batch_size=args.batch_size, shuffle=True)
test_loader = DataLoader(testset, batch_size=1)
else: # 生成随机数据
train_loader = [[torch.randn(size=(args.batch_size, args.times_num, args.bbox_input)),
torch.randn(size=(args.batch_size, 1)),
torch.randn(size=(args.batch_size, args.times_num, args.vel_input)),
torch.randn(size=(args.batch_size, args.times_num, args.bbox_input))]]
valid_loader = [[torch.randn(size=(args.batch_size, args.times_num, args.bbox_input)),
torch.randn(size=(args.batch_size, 1)),
torch.randn(size=(args.batch_size, args.times_num, args.vel_input)),
torch.randn(size=(args.batch_size, args.times_num, args.bbox_input))]]
test_loader = [[torch.randn(size=(args.batch_size, args.times_num, args.bbox_input)),
torch.randn(size=(args.batch_size, 1)),
torch.randn(size=(args.batch_size, args.times_num, args.vel_input)),
torch.randn(size=(args.batch_size, args.times_num, args.bbox_input))]]
print('Start Training Loop... \n')
model = Model(args)
model.to(device)
optimizer = torch.optim.Adam(model.parameters(), lr=args.lr, betas=(0.9, 0.98), eps=1e-6)
cls_criterion = nn.BCELoss()
reg_criterion = nn.MSELoss()
model_folder_name = args.set_path + '_' + args.bh
checkpoint_filepath = 'checkpoints/{}.pt'.format(model_folder_name)
writer = SummaryWriter('logs/{}'.format(model_folder_name))
train(model, train_loader, valid_loader, cls_criterion, reg_criterion, optimizer, checkpoint_filepath, writer, args=args)
#Test
model = Model(args)
model.to(device)
checkpoint = torch.load(checkpoint_filepath)
model.load_state_dict(checkpoint['model_state_dict'])
preds, labels = test(model, test_loader)
pred_cpu = torch.Tensor.cpu(preds)
label_cpu = torch.Tensor.cpu(labels)
acc = accuracy_score(label_cpu, np.round(pred_cpu))
f1 = f1_score(label_cpu, np.round(pred_cpu))
pre_s = precision_score(label_cpu, np.round(pred_cpu))
recall_s = recall_score(label_cpu, np.round(pred_cpu))
auc = roc_auc_score(label_cpu, np.round(pred_cpu))
matrix = confusion_matrix(label_cpu, np.round(pred_cpu))
print(f'Acc: {acc}\n f1: {f1}\n precision_score: {pre_s}\n recall_score: {recall_s}\n roc_auc_score: {auc}\n confusion_matrix: {matrix}')
if __name__ == '__main__':
torch.cuda.empty_cache()
parser = argparse.ArgumentParser('Pedestrain Crossing Intention Prediction.')
parser.add_argument('--epochs', type=int, default=2000, help='Number of epochs to train.')
parser.add_argument('--set_path', type=str, default='JAAD')
parser.add_argument('--bh', type=str, default='beh', help='all or beh, in JAAD dataset.')
parser.add_argument('--balance', type=bool, default=True, help='balance or not for test dataset.')
parser.add_argument('--seed', type=int, default=42)
parser.add_argument('--d_model', type=int, default=256, help='the dimension after embedding.')
parser.add_argument('--dff', type=int, default=512, help='the number of the units.')
parser.add_argument('--num_heads', type=int, default=8, help='number of the heads of the multi-head model.')
parser.add_argument('--bbox_input', type=int, default=4, help='dimension of bbox.')
parser.add_argument('--vel_input', type=int, default=1, help='dimension of velocity.')
parser.add_argument('--time_crop', type=bool, default=False)
parser.add_argument('--batch_size', type=int, default=64, help='size of batch.')
parser.add_argument('--lr', type=float, default=0.001, help='learning rate to train.')
parser.add_argument('--num_layers', type=int, default=4, help='the number of layers.')
parser.add_argument('--times_num', type=int, default=32, help='')
parser.add_argument('--num_bnks', type=int, default=9, help='')
parser.add_argument('--bnks_layers', type=int, default=9, help='')
parser.add_argument('--sta_f', type=int, default=8)
parser.add_argument('--end_f', type=int, default=12)
args = parser.parse_args()
main(args)