Spaces:
Running
on
Zero
Running
on
Zero
""" | |
A standalone PyTorch implementation for fast and efficient bicubic resampling. | |
The resulting values are the same to MATLAB function imresize('bicubic'). | |
## Author: Sanghyun Son | |
## Email: sonsang35@gmail.com (primary), thstkdgus35@snu.ac.kr (secondary) | |
## Version: 1.2.0 | |
## Last update: July 9th, 2020 (KST) | |
Dependency: torch | |
Example:: | |
import torch | |
import core | |
x = torch.arange(16).float().view(1, 1, 4, 4) | |
y = core.imresize(x, sizes=(3, 3)) | |
print(y) | |
tensor([[[[ 0.7506, 2.1004, 3.4503], | |
[ 6.1505, 7.5000, 8.8499], | |
[11.5497, 12.8996, 14.2494]]]]) | |
""" | |
import math | |
import typing | |
import torch | |
from torch.nn import functional as F | |
__all__ = ['imresize'] | |
_I = typing.Optional[int] | |
_D = typing.Optional[torch.dtype] | |
def nearest_contribution(x: torch.Tensor) -> torch.Tensor: | |
range_around_0 = torch.logical_and(x.gt(-0.5), x.le(0.5)) | |
cont = range_around_0.to(dtype=x.dtype) | |
return cont | |
def linear_contribution(x: torch.Tensor) -> torch.Tensor: | |
ax = x.abs() | |
range_01 = ax.le(1) | |
cont = (1 - ax) * range_01.to(dtype=x.dtype) | |
return cont | |
def cubic_contribution(x: torch.Tensor, a: float = -0.5) -> torch.Tensor: | |
ax = x.abs() | |
ax2 = ax * ax | |
ax3 = ax * ax2 | |
range_01 = ax.le(1) | |
range_12 = torch.logical_and(ax.gt(1), ax.le(2)) | |
cont_01 = (a + 2) * ax3 - (a + 3) * ax2 + 1 | |
cont_01 = cont_01 * range_01.to(dtype=x.dtype) | |
cont_12 = (a * ax3) - (5 * a * ax2) + (8 * a * ax) - (4 * a) | |
cont_12 = cont_12 * range_12.to(dtype=x.dtype) | |
cont = cont_01 + cont_12 | |
return cont | |
def gaussian_contribution(x: torch.Tensor, sigma: float = 2.0) -> torch.Tensor: | |
range_3sigma = (x.abs() <= 3 * sigma + 1) | |
# Normalization will be done after | |
cont = torch.exp(-x.pow(2) / (2 * sigma**2)) | |
cont = cont * range_3sigma.to(dtype=x.dtype) | |
return cont | |
def discrete_kernel(kernel: str, scale: float, antialiasing: bool = True) -> torch.Tensor: | |
''' | |
For downsampling with integer scale only. | |
''' | |
downsampling_factor = int(1 / scale) | |
if kernel == 'cubic': | |
kernel_size_orig = 4 | |
else: | |
raise ValueError('Pass!') | |
if antialiasing: | |
kernel_size = kernel_size_orig * downsampling_factor | |
else: | |
kernel_size = kernel_size_orig | |
if downsampling_factor % 2 == 0: | |
a = kernel_size_orig * (0.5 - 1 / (2 * kernel_size)) | |
else: | |
kernel_size -= 1 | |
a = kernel_size_orig * (0.5 - 1 / (kernel_size + 1)) | |
with torch.no_grad(): | |
r = torch.linspace(-a, a, steps=kernel_size) | |
k = cubic_contribution(r).view(-1, 1) | |
k = torch.matmul(k, k.t()) | |
k /= k.sum() | |
return k | |
def reflect_padding(x: torch.Tensor, dim: int, pad_pre: int, pad_post: int) -> torch.Tensor: | |
''' | |
Apply reflect padding to the given Tensor. | |
Note that it is slightly different from the PyTorch functional.pad, | |
where boundary elements are used only once. | |
Instead, we follow the MATLAB implementation | |
which uses boundary elements twice. | |
For example, | |
[a, b, c, d] would become [b, a, b, c, d, c] with the PyTorch implementation, | |
while our implementation yields [a, a, b, c, d, d]. | |
''' | |
b, c, h, w = x.size() | |
if dim == 2 or dim == -2: | |
padding_buffer = x.new_zeros(b, c, h + pad_pre + pad_post, w) | |
padding_buffer[..., pad_pre:(h + pad_pre), :].copy_(x) | |
for p in range(pad_pre): | |
padding_buffer[..., pad_pre - p - 1, :].copy_(x[..., p, :]) | |
for p in range(pad_post): | |
padding_buffer[..., h + pad_pre + p, :].copy_(x[..., -(p + 1), :]) | |
else: | |
padding_buffer = x.new_zeros(b, c, h, w + pad_pre + pad_post) | |
padding_buffer[..., pad_pre:(w + pad_pre)].copy_(x) | |
for p in range(pad_pre): | |
padding_buffer[..., pad_pre - p - 1].copy_(x[..., p]) | |
for p in range(pad_post): | |
padding_buffer[..., w + pad_pre + p].copy_(x[..., -(p + 1)]) | |
return padding_buffer | |
def padding(x: torch.Tensor, | |
dim: int, | |
pad_pre: int, | |
pad_post: int, | |
padding_type: typing.Optional[str] = 'reflect') -> torch.Tensor: | |
if padding_type is None: | |
return x | |
elif padding_type == 'reflect': | |
x_pad = reflect_padding(x, dim, pad_pre, pad_post) | |
else: | |
raise ValueError('{} padding is not supported!'.format(padding_type)) | |
return x_pad | |
def get_padding(base: torch.Tensor, kernel_size: int, x_size: int) -> typing.Tuple[int, int, torch.Tensor]: | |
base = base.long() | |
r_min = base.min() | |
r_max = base.max() + kernel_size - 1 | |
if r_min <= 0: | |
pad_pre = -r_min | |
pad_pre = pad_pre.item() | |
base += pad_pre | |
else: | |
pad_pre = 0 | |
if r_max >= x_size: | |
pad_post = r_max - x_size + 1 | |
pad_post = pad_post.item() | |
else: | |
pad_post = 0 | |
return pad_pre, pad_post, base | |
def get_weight(dist: torch.Tensor, | |
kernel_size: int, | |
kernel: str = 'cubic', | |
sigma: float = 2.0, | |
antialiasing_factor: float = 1) -> torch.Tensor: | |
buffer_pos = dist.new_zeros(kernel_size, len(dist)) | |
for idx, buffer_sub in enumerate(buffer_pos): | |
buffer_sub.copy_(dist - idx) | |
# Expand (downsampling) / Shrink (upsampling) the receptive field. | |
buffer_pos *= antialiasing_factor | |
if kernel == 'cubic': | |
weight = cubic_contribution(buffer_pos) | |
elif kernel == 'gaussian': | |
weight = gaussian_contribution(buffer_pos, sigma=sigma) | |
else: | |
raise ValueError('{} kernel is not supported!'.format(kernel)) | |
weight /= weight.sum(dim=0, keepdim=True) | |
return weight | |
def reshape_tensor(x: torch.Tensor, dim: int, kernel_size: int) -> torch.Tensor: | |
# Resize height | |
if dim == 2 or dim == -2: | |
k = (kernel_size, 1) | |
h_out = x.size(-2) - kernel_size + 1 | |
w_out = x.size(-1) | |
# Resize width | |
else: | |
k = (1, kernel_size) | |
h_out = x.size(-2) | |
w_out = x.size(-1) - kernel_size + 1 | |
unfold = F.unfold(x, k) | |
unfold = unfold.view(unfold.size(0), -1, h_out, w_out) | |
return unfold | |
def reshape_input(x: torch.Tensor) -> typing.Tuple[torch.Tensor, _I, _I, int, int]: | |
if x.dim() == 4: | |
b, c, h, w = x.size() | |
elif x.dim() == 3: | |
c, h, w = x.size() | |
b = None | |
elif x.dim() == 2: | |
h, w = x.size() | |
b = c = None | |
else: | |
raise ValueError('{}-dim Tensor is not supported!'.format(x.dim())) | |
x = x.view(-1, 1, h, w) | |
return x, b, c, h, w | |
def reshape_output(x: torch.Tensor, b: _I, c: _I) -> torch.Tensor: | |
rh = x.size(-2) | |
rw = x.size(-1) | |
# Back to the original dimension | |
if b is not None: | |
x = x.view(b, c, rh, rw) # 4-dim | |
else: | |
if c is not None: | |
x = x.view(c, rh, rw) # 3-dim | |
else: | |
x = x.view(rh, rw) # 2-dim | |
return x | |
def cast_input(x: torch.Tensor) -> typing.Tuple[torch.Tensor, _D]: | |
if x.dtype != torch.float32 or x.dtype != torch.float64: | |
dtype = x.dtype | |
x = x.float() | |
else: | |
dtype = None | |
return x, dtype | |
def cast_output(x: torch.Tensor, dtype: _D) -> torch.Tensor: | |
if dtype is not None: | |
if not dtype.is_floating_point: | |
x = x - x.detach() + x.round() | |
# To prevent over/underflow when converting types | |
if dtype is torch.uint8: | |
x = x.clamp(0, 255) | |
x = x.to(dtype=dtype) | |
return x | |
def resize_1d(x: torch.Tensor, | |
dim: int, | |
size: int, | |
scale: float, | |
kernel: str = 'cubic', | |
sigma: float = 2.0, | |
padding_type: str = 'reflect', | |
antialiasing: bool = True) -> torch.Tensor: | |
''' | |
Args: | |
x (torch.Tensor): A torch.Tensor of dimension (B x C, 1, H, W). | |
dim (int): | |
scale (float): | |
size (int): | |
Return: | |
''' | |
# Identity case | |
if scale == 1: | |
return x | |
# Default bicubic kernel with antialiasing (only when downsampling) | |
if kernel == 'cubic': | |
kernel_size = 4 | |
else: | |
kernel_size = math.floor(6 * sigma) | |
if antialiasing and (scale < 1): | |
antialiasing_factor = scale | |
kernel_size = math.ceil(kernel_size / antialiasing_factor) | |
else: | |
antialiasing_factor = 1 | |
# We allow margin to both sizes | |
kernel_size += 2 | |
# Weights only depend on the shape of input and output, | |
# so we do not calculate gradients here. | |
with torch.no_grad(): | |
pos = torch.linspace( | |
0, | |
size - 1, | |
steps=size, | |
dtype=x.dtype, | |
device=x.device, | |
) | |
pos = (pos + 0.5) / scale - 0.5 | |
base = pos.floor() - (kernel_size // 2) + 1 | |
dist = pos - base | |
weight = get_weight( | |
dist, | |
kernel_size, | |
kernel=kernel, | |
sigma=sigma, | |
antialiasing_factor=antialiasing_factor, | |
) | |
pad_pre, pad_post, base = get_padding(base, kernel_size, x.size(dim)) | |
# To backpropagate through x | |
x_pad = padding(x, dim, pad_pre, pad_post, padding_type=padding_type) | |
unfold = reshape_tensor(x_pad, dim, kernel_size) | |
# Subsampling first | |
if dim == 2 or dim == -2: | |
sample = unfold[..., base, :] | |
weight = weight.view(1, kernel_size, sample.size(2), 1) | |
else: | |
sample = unfold[..., base] | |
weight = weight.view(1, kernel_size, 1, sample.size(3)) | |
# Apply the kernel | |
x = sample * weight | |
x = x.sum(dim=1, keepdim=True) | |
return x | |
def downsampling_2d(x: torch.Tensor, k: torch.Tensor, scale: int, padding_type: str = 'reflect') -> torch.Tensor: | |
c = x.size(1) | |
k_h = k.size(-2) | |
k_w = k.size(-1) | |
k = k.to(dtype=x.dtype, device=x.device) | |
k = k.view(1, 1, k_h, k_w) | |
k = k.repeat(c, c, 1, 1) | |
e = torch.eye(c, dtype=k.dtype, device=k.device, requires_grad=False) | |
e = e.view(c, c, 1, 1) | |
k = k * e | |
pad_h = (k_h - scale) // 2 | |
pad_w = (k_w - scale) // 2 | |
x = padding(x, -2, pad_h, pad_h, padding_type=padding_type) | |
x = padding(x, -1, pad_w, pad_w, padding_type=padding_type) | |
y = F.conv2d(x, k, padding=0, stride=scale) | |
return y | |
def imresize(x: torch.Tensor, | |
scale: typing.Optional[float] = None, | |
sizes: typing.Optional[typing.Tuple[int, int]] = None, | |
kernel: typing.Union[str, torch.Tensor] = 'cubic', | |
sigma: float = 2, | |
rotation_degree: float = 0, | |
padding_type: str = 'reflect', | |
antialiasing: bool = True) -> torch.Tensor: | |
""" | |
Args: | |
x (torch.Tensor): | |
scale (float): | |
sizes (tuple(int, int)): | |
kernel (str, default='cubic'): | |
sigma (float, default=2): | |
rotation_degree (float, default=0): | |
padding_type (str, default='reflect'): | |
antialiasing (bool, default=True): | |
Return: | |
torch.Tensor: | |
""" | |
if scale is None and sizes is None: | |
raise ValueError('One of scale or sizes must be specified!') | |
if scale is not None and sizes is not None: | |
raise ValueError('Please specify scale or sizes to avoid conflict!') | |
x, b, c, h, w = reshape_input(x) | |
if sizes is None and scale is not None: | |
''' | |
# Check if we can apply the convolution algorithm | |
scale_inv = 1 / scale | |
if isinstance(kernel, str) and scale_inv.is_integer(): | |
kernel = discrete_kernel(kernel, scale, antialiasing=antialiasing) | |
elif isinstance(kernel, torch.Tensor) and not scale_inv.is_integer(): | |
raise ValueError( | |
'An integer downsampling factor ' | |
'should be used with a predefined kernel!' | |
) | |
''' | |
# Determine output size | |
sizes = (math.ceil(h * scale), math.ceil(w * scale)) | |
scales = (scale, scale) | |
if scale is None and sizes is not None: | |
scales = (sizes[0] / h, sizes[1] / w) | |
x, dtype = cast_input(x) | |
if isinstance(kernel, str) and sizes is not None: | |
# Core resizing module | |
x = resize_1d( | |
x, | |
-2, | |
size=sizes[0], | |
scale=scales[0], | |
kernel=kernel, | |
sigma=sigma, | |
padding_type=padding_type, | |
antialiasing=antialiasing) | |
x = resize_1d( | |
x, | |
-1, | |
size=sizes[1], | |
scale=scales[1], | |
kernel=kernel, | |
sigma=sigma, | |
padding_type=padding_type, | |
antialiasing=antialiasing) | |
elif isinstance(kernel, torch.Tensor) and scale is not None: | |
x = downsampling_2d(x, kernel, scale=int(1 / scale)) | |
x = reshape_output(x, b, c) | |
x = cast_output(x, dtype) | |
return x | |