model = EventDetector(pretrain=True,
width_mult=1.,
lstm_layers=1,
lstm_hidden=256,
device=device,
bidirectional=True,
dropout=False,
use_no_element=use_no_element
)
#print('model.py, class EventDetector()')
freeze_layers(k, model)
#print('utils.py, func freeze_laters()')
model.train()
model.to(device)
print('Loading Data')
# TODO: vid_dirのpathをかえる。stsqの動画を切り出したimage全部が含まれているdirにする
if use_no_element == False:
dataset = StsqDB(data_file='data/no_ele/seq_length_{}/train_split_{}.pkl'.format(args.seq_length, args.split),
vid_dir='/home/akiho/projects/golfdb/data/videos_40/',
seq_length=int(seq_length),
transform=transforms.Compose([ToTensor(),
Normalize([0.485, 0.456, 0.406], [0.229, 0.224, 0.225])]),
train=True)
else:
dataset = StsqDB(data_file='data/seq_length_{}/train_split_{}.pkl'.format(args.seq_length, args.split),
vid_dir='/home/akiho/projects/golfdb/data/videos_40/',
seq_length=int(seq_length),
# transform=transforms.Compose(
# [ToTensor(), Normalize([0.485, 0.456, 0.406], [0.229, 0.224, 0.225])]
# ),
train=True,
base_list=base_list,
event_th=event_th,
median=True,
)
model.cnn[0][0] = nn.Conv2d(len(base_list),
32,
3,
stride=2,
padding=1,
bias=False)
model = model.to(device)
# freeze_layers(k, model)
model.train()
# model.cuda()
data_loader = DataLoader(dataset,
batch_size=bs,
shuffle=True,
num_workers=n_cpu,
drop_last=True)
# the 8 golf swing events are classes 0 through 7, no-event is class 8
# the ratio of events to no-events is approximately 1:35 so weight classes accordingly:
weights = torch.FloatTensor(
[1 / 8, 1 / 8, 1 / 8, 1 / 8, 1 / 8, 1 / 8, 1 / 8, 1 / 8,
1 / 35]).to(device)