parser.add_argument('--no-cuda', action='store_true', help='Do not use cuda.')
parser.add_argument('--eval_input',
action='store_true',
help='Evaluate inputs instead.')
args = parser.parse_args()
cfg = config.load_config(args.config, 'configs/default.yaml')
is_cuda = (torch.cuda.is_available() and not args.no_cuda)
device = torch.device("cuda" if is_cuda else "cpu")
# Dataset
points_field = data.PointsField(
cfg['data']['points_iou_file'],
unpackbits=cfg['data']['points_unpackbits'],
)
pointcloud_field = data.PointCloudField(cfg['data']['pointcloud_chamfer_file'])
fields = {
'points_iou': points_field,
'pointcloud_chamfer': pointcloud_field,
'idx': data.IndexField(),
}
print('Test split: ', cfg['data']['test_split'])
dataset_folder = cfg['data']['path']
dataset = data.Shapes3dDataset_Human(dataset_folder,
fields,
cfg['data']['test_split'],
categories=cfg['data']['classes'])
# Evaluator
def get_inputs_field(mode, cfg, use_target_domain = False):
''' Returns the inputs fields.
Args:
mode (str): the mode which is used
cfg (dict): config dictionary
use_target_domain (bool): whether to use the target_domain dataset
'''
input_type = cfg['data']['input_type']
with_transforms = cfg['data']['with_transforms']
if input_type is None:
inputs_field = None
elif input_type == 'img':
if mode == 'train' and cfg['data']['img_augment']:
resize_op = transforms.RandomResizedCrop(
cfg['data']['img_size'], (0.75, 1.), (1., 1.))
else:
resize_op = transforms.Resize((cfg['data']['img_size']))
transform = transforms.Compose([
resize_op, transforms.ToTensor(),
])
with_camera = cfg['data']['img_with_camera']
if mode == 'train':
random_view = True
else:
random_view = False
if use_target_domain:
img_folder_name = cfg['data']['uda_img_folder']
filename_pattern= cfg['data']['uda_bg_configure']
else:
img_folder_name = cfg['data']['img_folder']
filename_pattern = cfg['data']['img_filename_pattern']
inputs_field = data.ImagesField(
img_folder_name, transform,
with_camera=with_camera, random_view=random_view, filename_pattern=filename_pattern, extensions=['jpg', 'jpeg', 'png']
)
elif input_type == 'pointcloud':
transform = transforms.Compose([
data.SubsamplePointcloud(cfg['data']['pointcloud_n']),
data.PointcloudNoise(cfg['data']['pointcloud_noise'])
])
with_transforms = cfg['data']['with_transforms']
inputs_field = data.PointCloudField(
cfg['data']['pointcloud_file'], transform,
with_transforms=with_transforms
)
elif input_type == 'voxels':
inputs_field = data.VoxelsField(
cfg['data']['voxels_file']
)
elif input_type == 'idx':
inputs_field = data.IndexField()
else:
raise ValueError(
'Invalid input type (%s)' % input_type)
return inputs_field
def get_inputs_field(mode, cfg):
''' Returns the inputs fields.
Args:
mode (str): the mode which is used
cfg (dict): config dictionary
'''
input_type = cfg['data']['input_type']
with_transforms = cfg['data']['with_transforms']
n_views = cfg['data']['n_views']
if input_type is None:
inputs_field = None
elif input_type == 'img':
if mode == 'train' and cfg['data']['img_augment']:
resize_op = transforms.RandomResizedCrop(cfg['data']['img_size'],
(0.75, 1.), (1., 1.))
else:
resize_op = transforms.Resize((cfg['data']['img_size']))
transform = transforms.Compose([
resize_op,
transforms.ToTensor(),
])
with_camera = cfg['data']['img_with_camera']
if mode == 'train':
random_view = True
else:
random_view = False
inputs_field = data.ImagesField(cfg['data']['img_folder'],
transform,
with_camera=with_camera,
random_view=random_view)
elif input_type == 'multi_img':
if mode == 'train' and cfg['data']['img_augment']:
resize_op = transforms.RandomResizedCrop(cfg['data']['img_size'],
(0.75, 1.), (1., 1.))
else:
resize_op = transforms.Resize((cfg['data']['img_size']))
transform = transforms.Compose([
resize_op,
transforms.ToTensor(),
])
with_camera = cfg['data']['img_with_camera']
if mode == 'train':
random_view = True
else:
random_view = False
inputs_field = data.MultiImageField(cfg['data']['img_folder'],
transform=transform,
with_camera=with_camera,
random_view=random_view,
n_views=n_views)
elif input_type == 'pointcloud':
transform = transforms.Compose([
data.SubsamplePointcloud(cfg['data']['pointcloud_n']),
data.PointcloudNoise(cfg['data']['pointcloud_noise'])
])
with_transforms = cfg['data']['with_transforms']
inputs_field = data.PointCloudField(cfg['data']['pointcloud_file'],
transform,
with_transforms=with_transforms)
elif input_type == 'voxels':
inputs_field = data.VoxelsField(cfg['data']['voxels_file'])
elif input_type == 'idx':
inputs_field = data.IndexField()
else:
raise ValueError('Invalid input type (%s)' % input_type)
return inputs_field
def get_data_fields(mode, cfg):
''' Returns the data fields.
Args:
mode (str): the mode which is used
cfg (dict): imported yaml config
'''
points_transform = data.SubsamplePoints(cfg['data']['points_subsample'])
if cfg.get('sdf_generation', False):
points_transform = None
with_transforms = cfg['model']['use_camera']
fields = {}
fields['points'] = data.PointsField(
cfg['data']['points_file'],
points_transform,
with_transforms=with_transforms,
unpackbits=cfg['data']['points_unpackbits'],
)
if not cfg.get('sdf_generation', False) and cfg['trainer'].get(
'is_sdf', False):
sdf_points_transform = data.SubsampleSDFPoints(
cfg['data']['points_subsample'])
fields['sdf_points'] = data.SDFPointsField(
cfg['data']['sdf_points_file'],
sdf_points_transform,
with_transforms=with_transforms)
pointcloud_transform = data.SubsamplePointcloud(
cfg['data']['pointcloud_target_n'])
if cfg.get('sdf_generation', False):
pointcloud_transform = None
fields['pointcloud'] = data.PointCloudField(cfg['data']['pointcloud_file'],
pointcloud_transform,
with_transforms=True)
fields['angles'] = nsd_data.SphericalCoordinateField(
cfg['data']['primitive_points_sample_n'],
mode,
is_normal_icosahedron=cfg['data'].get('is_normal_icosahedron', False),
is_normal_uv_sphere=cfg['data'].get('is_normal_uv_sphere', False),
icosahedron_subdiv=cfg['data'].get('icosahedron_subdiv', 2),
icosahedron_uv_margin=cfg['data'].get('icosahedron_uv_margin', 1e-5),
icosahedron_uv_margin_phi=cfg['data'].get('icosahedron_uv_margin_phi',
1e-5),
uv_sphere_length=cfg['data'].get('uv_sphere_length', 20),
normal_mesh_no_invert=cfg['data'].get('normal_mesh_no_invert', False))
if mode in ('val', 'test'):
points_iou_file = cfg['data']['points_iou_file']
voxels_file = cfg['data']['voxels_file']
if points_iou_file is not None:
fields['points_iou'] = data.PointsField(
points_iou_file,
with_transforms=with_transforms,
unpackbits=cfg['data']['points_unpackbits'],
)
if voxels_file is not None:
fields['voxels'] = data.VoxelsField(voxels_file)
return fields
def get_inputs_field(mode, cfg):
""" Returns the inputs fields.
Args:
mode (str): the mode which is used
cfg (dict): config dictionary
"""
input_type = cfg["data"]["input_type"]
with_transforms = cfg["data"]["with_transforms"]
if input_type is None:
inputs_field = None
elif input_type == "img":
transform = None
if mode == "train" and cfg["data"]["img_augment"]:
# resize_op = transforms.RandomResizedCrop(cfg["data"]["img_size"],(0.75, 1.0), (1.0, 1.0))
def preprocess(image):
# image = tf.image.crop_and_resize(
# image, crop_size=cfg["data"]["img_size"]) # CHECK
image = tf.image.resize(
image, [cfg["data"]["img_size"], cfg["data"]["img_size"]])
image /= 255.0
return image
transform = preprocess
else:
def preprocess(image):
# image = image[tf.newaxis, ...]
image = tf.image.resize(
image, [cfg["data"]["img_size"], cfg["data"]["img_size"]])
image /= 255.0
return image
transform = preprocess
# transform = transforms.Compose([
# resize_op,
# transforms.ToTensor(),
# ])
with_camera = cfg["data"]["img_with_camera"]
if mode == "train":
random_view = True
else:
random_view = False
inputs_field = data.ImagesField(
cfg["data"]["img_folder"],
transform,
with_camera=with_camera,
random_view=random_view,
)
elif input_type == "pointcloud":
def preprocess(points):
output = data.SubsamplePointcloud(
cfg["data"]["pointcloud_n"])(points)
output = data.PointcloudNoise(
cfg["data"]["pointcloud_noise"])(output)
return output
transform = preprocess
with_transforms = cfg["data"]["with_transforms"]
inputs_field = data.PointCloudField(cfg["data"]["pointcloud_file"],
transform,
with_transforms=with_transforms)
elif input_type == "voxels":
inputs_field = data.VoxelsField(cfg["data"]["voxels_file"])
elif input_type == "idx":
inputs_field = data.IndexField()
else:
raise ValueError("Invalid input type (%s)" % input_type)
return inputs_field