# generator
model = FullNetwork(vp_value_count=VP_VALUE_COUNT, stdev=STDEV,
output_shape=(BATCH_SIZE, CHANNELS, FRAMES, HEIGHT, WIDTH))
model = model.to(device)
if device == 'cuda':
net = torch.nn.DataParallel(model)
cudnn.benchmark = True
criterion = nn.MSELoss()
optimizer = optim.Adam(model.parameters(), lr=LR)
# data
trainset = NTUDataset(root_dir=data_root_dir, data_file=train_split, param_file=param_file,
resize_height=HEIGHT, resize_width=WIDTH,
clip_len=FRAMES, skip_len=SKIP_LEN,
random_all=RANDOM_ALL, precrop=PRECROP)
trainloader = torch.utils.data.DataLoader(trainset, batch_size=BATCH_SIZE, shuffle=True, num_workers=2)
testset = NTUDataset(root_dir=data_root_dir, data_file=test_split, param_file=param_file,
resize_height=HEIGHT, resize_width=WIDTH,
clip_len=FRAMES, skip_len=SKIP_LEN,
random_all=RANDOM_ALL, precrop=PRECROP)
testloader = torch.utils.data.DataLoader(testset, batch_size=BATCH_SIZE, shuffle=False, num_workers=2)
elif DATASET.lower() == 'panoptic':
data_root_dir, train_split, test_split, close_cams_file, weight_file = panoptic_config()
# generator
model = FullNetwork(vp_value_count=VP_VALUE_COUNT, stdev=STDEV,
output_shape=(BATCH_SIZE, CHANNELS, FRAMES, HEIGHT, WIDTH), use_est_vp=False)
#
# for sample in data_file:
# sample = sample.split(' ')
# sample_id = sample[0][sample[0].index('/') + 1:]
# scene, pid, rid, action = decrypt_vid_name(sample_id)
# print(action)
# if action == 18:
# if pid not in actors:
# actors.append(pid)
#
# actors.sort()
#
# print(actors)
trainset = NTUDataset(root_dir=data_root_dir, data_file=train_split, param_file=param_file,
resize_height=HEIGHT, resize_width=WIDTH,
clip_len=FRAMES, skip_len=SKIP_LEN,
random_all=RANDOM_ALL, precrop=PRECROP, diff_actors=DIFF_ACTORS, diff_scenes=DIFF_SCENES)
trainloader = torch.utils.data.DataLoader(trainset, batch_size=BATCH_SIZE, shuffle=True, num_workers=2)
for batch_idx, info in enumerate(trainloader):
if info[0] != info[1]:
print('cry')
# print(info)
# import torch
# x = torch.tensor([2.0,2.0])
# print(x.requires_grad_())
# x = x.requires_grad_()
