def __call__(self, file_a, file_b, label, mask):
angle = float(
torch.empty(1).uniform_(float(self.degrees[0]),
float(self.degrees[1])).item())
# print('angle: ', angle)
center_f = [0.0, 0.0]
matrix = tf._get_inverse_affine_matrix(center_f, -angle, [0.0, 0.0],
1.0, [0.0, 0.0])
return (F_t.rotate(file_a,
matrix=matrix,
resample=self.resample,
expand=self.expand,
fill=self.fill),
F_t.rotate(file_b,
matrix=matrix,
resample=self.resample,
expand=self.expand,
fill=self.fill),
F_t.rotate(label.unsqueeze(0),
matrix=matrix,
resample=self.resample,
expand=self.expand,
fill=self.fill).squeeze(0),
F_t.rotate(mask.unsqueeze(0),
matrix=matrix,
resample=self.resample,
expand=self.expand,
fill=self.fill).squeeze(0))
def rotate_image_tensor(
img: torch.Tensor,
angle: float,
interpolation: InterpolationMode = InterpolationMode.NEAREST,
expand: bool = False,
fill: Optional[List[float]] = None,
center: Optional[List[float]] = None,
) -> torch.Tensor:
center_f = [0.0, 0.0]
if center is not None:
if expand:
warnings.warn(
"The provided center argument has no effect on the result if expand is True"
)
else:
_, height, width = get_dimensions_image_tensor(img)
# Center values should be in pixel coordinates but translated such that (0, 0) corresponds to image center.
center_f = [
1.0 * (c - s * 0.5) for c, s in zip(center, [width, height])
]
# due to current incoherence of rotation angle direction between affine and rotate implementations
# we need to set -angle.
matrix = _get_inverse_affine_matrix(center_f, -angle, [0.0, 0.0], 1.0,
[0.0, 0.0])
return _FT.rotate(img,
matrix,
interpolation=interpolation.value,
expand=expand,
fill=fill)
def __call__(self, file):
file = torch.from_numpy(file)
angle = float(torch.empty(1).uniform_(float(self.degrees[0]), float(self.degrees[1])).item())
# print('angle: ', angle)
center_f = [0.0, 0.0]
matrix = tf._get_inverse_affine_matrix(center_f, -angle, [0.0, 0.0], 1.0, [0.0, 0.0])
return F_t.rotate(file, matrix=matrix, resample=self.resample,
expand=self.expand, fill=self.fill)