def create_train_val_dataloader(configs):
"""Create dataloader for training and validate"""
train_aug_transforms = OneOf([
Random_Rotation(limit_angle=20., p=1.0),
Random_Scaling(scaling_range=(0.95, 1.05), p=1.0)
],
p=0.6)
train_dataset = KittiDataset(configs.dataset_dir,
split='train',
mode='train',
aug_transforms=train_aug_transforms,
hflip_prob=0.5,
multiscale=configs.multiscale_training,
num_samples=configs.num_samples,
mosaic=configs.mosaic,
random_padding=configs.random_padding)
train_sampler = None
if configs.distributed:
train_sampler = torch.utils.data.distributed.DistributedSampler(
train_dataset)
train_dataloader = DataLoader(train_dataset,
batch_size=configs.batch_size,
shuffle=(train_sampler is None),
pin_memory=configs.pin_memory,
num_workers=configs.num_workers,
sampler=train_sampler,
collate_fn=train_dataset.collate_fn)
val_sampler = None
val_dataset = KittiDataset(configs.dataset_dir,
split='val',
mode='val',
aug_transforms=None,
hflip_prob=0.,
multiscale=False,
num_samples=configs.num_samples,
mosaic=False,
random_padding=False)
if configs.distributed:
val_sampler = torch.utils.data.distributed.DistributedSampler(
val_dataset, shuffle=False)
val_dataloader = DataLoader(val_dataset,
batch_size=configs.batch_size,
shuffle=False,
pin_memory=configs.pin_memory,
num_workers=configs.num_workers,
sampler=val_sampler,
collate_fn=val_dataset.collate_fn)
return train_dataloader, val_dataloader, train_sampler
def create_train_dataloader(configs):
"""Create dataloader for training"""
train_lidar_aug = OneOf([
Random_Rotation(limit_angle=np.pi / 4, p=1.0),
Random_Scaling(scaling_range=(0.95, 1.05), p=1.0),
Random_Rotate_Individual_Box(limit_angle=np.pi / 10, p=1.0)
],
p=0.75)
train_dataset = KittiDataset(configs,
mode='train',
lidar_aug=train_lidar_aug,
aug_transforms=None,
num_samples=configs.num_samples)
train_sampler = None
if configs.distributed:
train_sampler = torch.utils.data.distributed.DistributedSampler(
train_dataset)
train_dataloader = DataLoader(train_dataset,
batch_size=configs.batch_size,
shuffle=(train_sampler is None),
pin_memory=configs.pin_memory,
num_workers=configs.num_workers,
sampler=train_sampler,
collate_fn=train_dataset.collate_fn)
return train_dataloader, train_sampler
def create_val_dataloader(configs):
"""Create dataloader for validation"""
val_sampler = None
val_dataset = KittiDataset(configs.dataset_dir,
split='val',
mode='val',
aug_transforms=None,
hflip_prob=0.,
multiscale=False,
num_samples=configs.num_samples,
mosaic=False,
random_padding=False)
if configs.distributed:
val_sampler = torch.utils.data.distributed.DistributedSampler(
val_dataset, shuffle=False)
val_dataloader = DataLoader(val_dataset,
batch_size=configs.batch_size,
shuffle=False,
pin_memory=configs.pin_memory,
num_workers=configs.num_workers,
sampler=val_sampler,
collate_fn=val_dataset.collate_fn)
return val_dataloader
def create_train_dataloader(configs):
"""Create dataloader for training"""
train_lidar_transforms = OneOf([
Random_Rotation(limit_angle=20., p=1.0),
Random_Scaling(scaling_range=(0.95, 1.05), p=1.0)
], p=0.66)
train_aug_transforms = Compose([
Horizontal_Flip(p=configs.hflip_prob),
Cutout(n_holes=configs.cutout_nholes, ratio=configs.cutout_ratio, fill_value=configs.cutout_fill_value,
p=configs.cutout_prob)
], p=1.)
train_dataset = KittiDataset(configs.dataset_dir, mode='train', lidar_transforms=train_lidar_transforms,
aug_transforms=train_aug_transforms, multiscale=configs.multiscale_training,
num_samples=configs.num_samples, mosaic=configs.mosaic,
random_padding=configs.random_padding)
train_sampler = None
if configs.distributed:
train_sampler = torch.utils.data.distributed.DistributedSampler(train_dataset)
train_dataloader = DataLoader(train_dataset, batch_size=configs.batch_size, shuffle=(train_sampler is None),
pin_memory=configs.pin_memory, num_workers=configs.num_workers, sampler=train_sampler,
collate_fn=train_dataset.collate_fn)
return train_dataloader, train_sampler
def create_train_dataloader(configs, voxel_generator):
"""Create dataloader for training"""
train_lidar_aug = OneOf([
Random_Rotation(limit_angle=np.pi / 4, p=1.0),
Random_Scaling(scaling_range=(0.95, 1.05), p=1.0),
],
p=0.66)
train_dataset = KittiDataset(configs,
voxel_generator,
mode='train',
lidar_aug=train_lidar_aug,
hflip_prob=configs.hflip_prob,
num_samples=configs.num_samples)
train_sampler = None
if configs.distributed:
train_sampler = torch.utils.data.distributed.DistributedSampler(
train_dataset)
train_dataloader = DataLoader(train_dataset,
batch_size=configs.batch_size,
shuffle=(train_sampler is None),
pin_memory=configs.pin_memory,
num_workers=configs.num_workers,
sampler=train_sampler,
collate_fn=merge_batch)
return train_dataloader, train_sampler
def create_anotha_dataloader(configs):
"""Create dataloader for testing on training dataset"""
print(configs.dataset_dir)
test_dataset = KittiDataset(configs.dataset_dir,
mode='test',
lidar_transforms=None,
aug_transforms=None,
multiscale=False,
num_samples=configs.num_samples,
mosaic=False,
random_padding=False,
switcheroo=True)
test_sampler = None
if configs.distributed:
test_sampler = torch.utils.data.distributed.DistributedSampler(
test_dataset)
test_dataloader = DataLoader(test_dataset,
batch_size=configs.batch_size,
shuffle=False,
pin_memory=configs.pin_memory,
num_workers=configs.num_workers,
sampler=test_sampler)
return test_dataloader
def create_val_dataloader(configs):
"""Create dataloader for validation"""
val_sampler = None
val_dataset = KittiDataset(configs, mode='val', lidar_aug=None, hflip_prob=0., num_samples=configs.num_samples)
if configs.distributed:
val_sampler = torch.utils.data.distributed.DistributedSampler(val_dataset, shuffle=False)
val_dataloader = DataLoader(val_dataset, batch_size=configs.batch_size, shuffle=False,
pin_memory=configs.pin_memory, num_workers=configs.num_workers, sampler=val_sampler)
return val_dataloader
def create_test_dataset(configs):
"""Create dataloader for testing phase"""
test_dataset = KittiDataset(configs.dataset_dir,
mode='test',
lidar_transforms=None,
aug_transforms=None,
multiscale=False,
num_samples=configs.num_samples,
mosaic=False,
random_padding=False)
return test_dataset
def create_test_dataloader(configs):
"""Create dataloader for testing phase"""
test_dataset = KittiDataset(configs, mode='test', lidar_aug=None, hflip_prob=0., num_samples=configs.num_samples)
test_sampler = None
if configs.distributed:
test_sampler = torch.utils.data.distributed.DistributedSampler(test_dataset)
test_dataloader = DataLoader(test_dataset, batch_size=configs.batch_size, shuffle=False,
pin_memory=configs.pin_memory, num_workers=configs.num_workers, sampler=test_sampler)
return test_dataloader
def create_train_val_dataloader(configs):
"""Create dataloader for training and validate"""
train_dataset = KittiDataset(configs.dataset_dir, split='train', mode='train', data_aug=True,
multiscale=configs.multiscale_training, num_samples=configs.num_samples)
train_sampler = None
if configs.distributed:
train_sampler = torch.utils.data.distributed.DistributedSampler(train_dataset)
train_dataloader = DataLoader(train_dataset, batch_size=configs.batch_size, shuffle=(train_sampler is None),
pin_memory=configs.pin_memory, num_workers=configs.num_workers, sampler=train_sampler,
collate_fn=train_dataset.collate_fn)
val_sampler = None
val_dataset = KittiDataset(configs.dataset_dir, split='val', mode='val', data_aug=False, multiscale=False,
num_samples=configs.num_samples)
if configs.distributed:
val_sampler = torch.utils.data.distributed.DistributedSampler(val_dataset, shuffle=False)
val_dataloader = DataLoader(val_dataset, batch_size=configs.batch_size, shuffle=False,
pin_memory=configs.pin_memory, num_workers=configs.num_workers, sampler=val_sampler,
collate_fn=val_dataset.collate_fn)
return train_dataloader, val_dataloader, train_sampler
def create_val_dataloader(configs):
"""Create dataloader for validation"""
val_aug_transforms = Compose([
Horizontal_Flip(p=configs.hflip_prob),
Cutout(n_holes=configs.cutout_nholes, ratio=configs.cutout_ratio, fill_value=configs.cutout_fill_value,
p=configs.cutout_prob)
], p=1.)
val_sampler = None
val_dataset = KittiDataset(configs.dataset_dir, mode='val', lidar_transforms=None,
aug_transforms=val_aug_transforms, multiscale=False, num_samples=configs.num_samples,
mosaic=False, random_padding=False)
if configs.distributed:
val_sampler = torch.utils.data.distributed.DistributedSampler(val_dataset, shuffle=False)
val_dataloader = DataLoader(val_dataset, batch_size=configs.batch_size, shuffle=False,
pin_memory=configs.pin_memory, num_workers=configs.num_workers, sampler=val_sampler,
collate_fn=val_dataset.collate_fn)
return val_dataloader
def create_test_dataloader(configs):
"""Create dataloader for testing phase"""
test_dataset = KittiDataset(configs.dataset_dir,
mode='test',
aug_transforms=None,
hflip_prob=0.,
multiscale=False,
num_samples=configs.num_samples,
mosaic=False,
random_padding=False)
test_sampler = None
if configs.distributed:
test_sampler = torch.utils.data.distributed.DistributedSampler(
test_dataset)
test_dataloader = DataLoader(test_dataset,
batch_size=configs.batch_size,
shuffle=False,
pin_memory=configs.pin_memory,
num_workers=configs.num_workers,
sampler=test_sampler)
return test_dataloader
def create_train_dataloader(configs):
"""Create dataloader for training"""
train_aug_transforms = album.Compose(
[album.RandomBrightnessContrast(p=0.5),
album.GaussNoise(p=0.5)], p=1.)
train_dataset = KittiDataset(configs,
mode='train',
aug_transforms=train_aug_transforms,
hflip_prob=configs.hflip_prob,
use_left_cam_prob=configs.use_left_cam_prob,
num_samples=configs.num_samples)
train_sampler = None
if configs.distributed:
train_sampler = torch.utils.data.distributed.DistributedSampler(
train_dataset)
train_dataloader = DataLoader(train_dataset,
batch_size=configs.batch_size,
shuffle=(train_sampler is None),
pin_memory=configs.pin_memory,
num_workers=configs.num_workers,
sampler=train_sampler)
return train_dataloader, train_sampler