functional.py 52.5 KB
Newer Older
1
import math
2
3
import numbers
import warnings
4
from enum import Enum
5
6

import numpy as np
vfdev's avatar
vfdev committed
7
from PIL import Image
8
9
10

import torch
from torch import Tensor
11
from typing import List, Tuple, Any, Optional
12

13
14
15
16
17
try:
    import accimage
except ImportError:
    accimage = None

18
19
20
from . import functional_pil as F_pil
from . import functional_tensor as F_t

21

22
class InterpolationMode(Enum):
23
24
25
26
27
28
29
30
31
32
33
34
    """Interpolation modes
    """
    NEAREST = "nearest"
    BILINEAR = "bilinear"
    BICUBIC = "bicubic"
    # For PIL compatibility
    BOX = "box"
    HAMMING = "hamming"
    LANCZOS = "lanczos"


# TODO: Once torchscript supports Enums with staticmethod
35
36
# this can be put into InterpolationMode as staticmethod
def _interpolation_modes_from_int(i: int) -> InterpolationMode:
37
    inverse_modes_mapping = {
38
39
40
41
42
43
        0: InterpolationMode.NEAREST,
        2: InterpolationMode.BILINEAR,
        3: InterpolationMode.BICUBIC,
        4: InterpolationMode.BOX,
        5: InterpolationMode.HAMMING,
        1: InterpolationMode.LANCZOS,
44
45
46
47
48
    }
    return inverse_modes_mapping[i]


pil_modes_mapping = {
49
50
51
52
53
54
    InterpolationMode.NEAREST: 0,
    InterpolationMode.BILINEAR: 2,
    InterpolationMode.BICUBIC: 3,
    InterpolationMode.BOX: 4,
    InterpolationMode.HAMMING: 5,
    InterpolationMode.LANCZOS: 1,
55
56
}

vfdev's avatar
vfdev committed
57
_is_pil_image = F_pil._is_pil_image
vfdev's avatar
vfdev committed
58
_parse_fill = F_pil._parse_fill
vfdev's avatar
vfdev committed
59
60
61


def _get_image_size(img: Tensor) -> List[int]:
62
    """Returns image size as [w, h]
vfdev's avatar
vfdev committed
63
64
65
    """
    if isinstance(img, torch.Tensor):
        return F_t._get_image_size(img)
66

vfdev's avatar
vfdev committed
67
    return F_pil._get_image_size(img)
68

vfdev's avatar
vfdev committed
69

70
def _get_image_num_channels(img: Tensor) -> int:
71
72
    """Returns number of image channels
    """
73
74
75
76
77
78
    if isinstance(img, torch.Tensor):
        return F_t._get_image_num_channels(img)

    return F_pil._get_image_num_channels(img)


vfdev's avatar
vfdev committed
79
80
@torch.jit.unused
def _is_numpy(img: Any) -> bool:
81
82
83
    return isinstance(img, np.ndarray)


vfdev's avatar
vfdev committed
84
85
@torch.jit.unused
def _is_numpy_image(img: Any) -> bool:
86
    return img.ndim in {2, 3}
87
88
89
90


def to_tensor(pic):
    """Convert a ``PIL Image`` or ``numpy.ndarray`` to tensor.
91
    This function does not support torchscript.
92

93
    See :class:`~torchvision.transforms.ToTensor` for more details.
94
95
96
97
98
99
100

    Args:
        pic (PIL Image or numpy.ndarray): Image to be converted to tensor.

    Returns:
        Tensor: Converted image.
    """
vfdev's avatar
vfdev committed
101
    if not(F_pil._is_pil_image(pic) or _is_numpy(pic)):
102
103
        raise TypeError('pic should be PIL Image or ndarray. Got {}'.format(type(pic)))

104
105
106
    if _is_numpy(pic) and not _is_numpy_image(pic):
        raise ValueError('pic should be 2/3 dimensional. Got {} dimensions.'.format(pic.ndim))

107
108
    if isinstance(pic, np.ndarray):
        # handle numpy array
surgan12's avatar
surgan12 committed
109
110
111
        if pic.ndim == 2:
            pic = pic[:, :, None]

112
        img = torch.from_numpy(pic.transpose((2, 0, 1))).contiguous()
113
        # backward compatibility
114
115
116
117
        if isinstance(img, torch.ByteTensor):
            return img.float().div(255)
        else:
            return img
118
119
120
121
122
123
124
125
126
127
128

    if accimage is not None and isinstance(pic, accimage.Image):
        nppic = np.zeros([pic.channels, pic.height, pic.width], dtype=np.float32)
        pic.copyto(nppic)
        return torch.from_numpy(nppic)

    # handle PIL Image
    if pic.mode == 'I':
        img = torch.from_numpy(np.array(pic, np.int32, copy=False))
    elif pic.mode == 'I;16':
        img = torch.from_numpy(np.array(pic, np.int16, copy=False))
129
130
    elif pic.mode == 'F':
        img = torch.from_numpy(np.array(pic, np.float32, copy=False))
131
132
    elif pic.mode == '1':
        img = 255 * torch.from_numpy(np.array(pic, np.uint8, copy=False))
133
134
    else:
        img = torch.ByteTensor(torch.ByteStorage.from_buffer(pic.tobytes()))
135
136

    img = img.view(pic.size[1], pic.size[0], len(pic.getbands()))
137
    # put it from HWC to CHW format
138
    img = img.permute((2, 0, 1)).contiguous()
139
140
141
142
143
144
    if isinstance(img, torch.ByteTensor):
        return img.float().div(255)
    else:
        return img


145
146
def pil_to_tensor(pic):
    """Convert a ``PIL Image`` to a tensor of the same type.
147
    This function does not support torchscript.
148

vfdev's avatar
vfdev committed
149
    See :class:`~torchvision.transforms.PILToTensor` for more details.
150
151
152
153
154
155
156

    Args:
        pic (PIL Image): Image to be converted to tensor.

    Returns:
        Tensor: Converted image.
    """
157
    if not F_pil._is_pil_image(pic):
158
159
160
        raise TypeError('pic should be PIL Image. Got {}'.format(type(pic)))

    if accimage is not None and isinstance(pic, accimage.Image):
161
162
        # accimage format is always uint8 internally, so always return uint8 here
        nppic = np.zeros([pic.channels, pic.height, pic.width], dtype=np.uint8)
163
164
165
166
167
168
169
170
171
172
173
        pic.copyto(nppic)
        return torch.as_tensor(nppic)

    # handle PIL Image
    img = torch.as_tensor(np.asarray(pic))
    img = img.view(pic.size[1], pic.size[0], len(pic.getbands()))
    # put it from HWC to CHW format
    img = img.permute((2, 0, 1))
    return img


174
175
def convert_image_dtype(image: torch.Tensor, dtype: torch.dtype = torch.float) -> torch.Tensor:
    """Convert a tensor image to the given ``dtype`` and scale the values accordingly
176
    This function does not support PIL Image.
177
178
179
180
181
182

    Args:
        image (torch.Tensor): Image to be converted
        dtype (torch.dtype): Desired data type of the output

    Returns:
vfdev's avatar
vfdev committed
183
        Tensor: Converted image
184
185
186
187
188
189
190
191
192
193
194
195

    .. note::

        When converting from a smaller to a larger integer ``dtype`` the maximum values are **not** mapped exactly.
        If converted back and forth, this mismatch has no effect.

    Raises:
        RuntimeError: When trying to cast :class:`torch.float32` to :class:`torch.int32` or :class:`torch.int64` as
            well as for trying to cast :class:`torch.float64` to :class:`torch.int64`. These conversions might lead to
            overflow errors since the floating point ``dtype`` cannot store consecutive integers over the whole range
            of the integer ``dtype``.
    """
196
197
198
199
    if not isinstance(image, torch.Tensor):
        raise TypeError('Input img should be Tensor Image')

    return F_t.convert_image_dtype(image, dtype)
200
201


202
def to_pil_image(pic, mode=None):
203
    """Convert a tensor or an ndarray to PIL Image. This function does not support torchscript.
204

205
    See :class:`~torchvision.transforms.ToPILImage` for more details.
206
207
208
209
210

    Args:
        pic (Tensor or numpy.ndarray): Image to be converted to PIL Image.
        mode (`PIL.Image mode`_): color space and pixel depth of input data (optional).

211
    .. _PIL.Image mode: https://pillow.readthedocs.io/en/latest/handbook/concepts.html#concept-modes
212
213
214
215

    Returns:
        PIL Image: Image converted to PIL Image.
    """
Varun Agrawal's avatar
Varun Agrawal committed
216
    if not(isinstance(pic, torch.Tensor) or isinstance(pic, np.ndarray)):
217
218
        raise TypeError('pic should be Tensor or ndarray. Got {}.'.format(type(pic)))

Varun Agrawal's avatar
Varun Agrawal committed
219
220
221
222
223
224
    elif isinstance(pic, torch.Tensor):
        if pic.ndimension() not in {2, 3}:
            raise ValueError('pic should be 2/3 dimensional. Got {} dimensions.'.format(pic.ndimension()))

        elif pic.ndimension() == 2:
            # if 2D image, add channel dimension (CHW)
Surgan Jandial's avatar
Surgan Jandial committed
225
            pic = pic.unsqueeze(0)
Varun Agrawal's avatar
Varun Agrawal committed
226

227
228
229
230
        # check number of channels
        if pic.shape[-3] > 4:
            raise ValueError('pic should not have > 4 channels. Got {} channels.'.format(pic.shape[-3]))

Varun Agrawal's avatar
Varun Agrawal committed
231
232
233
234
235
236
237
238
    elif isinstance(pic, np.ndarray):
        if pic.ndim not in {2, 3}:
            raise ValueError('pic should be 2/3 dimensional. Got {} dimensions.'.format(pic.ndim))

        elif pic.ndim == 2:
            # if 2D image, add channel dimension (HWC)
            pic = np.expand_dims(pic, 2)

239
240
241
242
        # check number of channels
        if pic.shape[-1] > 4:
            raise ValueError('pic should not have > 4 channels. Got {} channels.'.format(pic.shape[-1]))

243
    npimg = pic
Varun Agrawal's avatar
Varun Agrawal committed
244
    if isinstance(pic, torch.Tensor):
245
246
247
        if pic.is_floating_point() and mode != 'F':
            pic = pic.mul(255).byte()
        npimg = np.transpose(pic.cpu().numpy(), (1, 2, 0))
248
249
250
251
252
253
254
255
256
257

    if not isinstance(npimg, np.ndarray):
        raise TypeError('Input pic must be a torch.Tensor or NumPy ndarray, ' +
                        'not {}'.format(type(npimg)))

    if npimg.shape[2] == 1:
        expected_mode = None
        npimg = npimg[:, :, 0]
        if npimg.dtype == np.uint8:
            expected_mode = 'L'
vfdev's avatar
vfdev committed
258
        elif npimg.dtype == np.int16:
259
            expected_mode = 'I;16'
vfdev's avatar
vfdev committed
260
        elif npimg.dtype == np.int32:
261
262
263
264
265
266
267
268
            expected_mode = 'I'
        elif npimg.dtype == np.float32:
            expected_mode = 'F'
        if mode is not None and mode != expected_mode:
            raise ValueError("Incorrect mode ({}) supplied for input type {}. Should be {}"
                             .format(mode, np.dtype, expected_mode))
        mode = expected_mode

surgan12's avatar
surgan12 committed
269
270
271
272
273
274
275
276
    elif npimg.shape[2] == 2:
        permitted_2_channel_modes = ['LA']
        if mode is not None and mode not in permitted_2_channel_modes:
            raise ValueError("Only modes {} are supported for 2D inputs".format(permitted_2_channel_modes))

        if mode is None and npimg.dtype == np.uint8:
            mode = 'LA'

277
    elif npimg.shape[2] == 4:
surgan12's avatar
surgan12 committed
278
        permitted_4_channel_modes = ['RGBA', 'CMYK', 'RGBX']
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
        if mode is not None and mode not in permitted_4_channel_modes:
            raise ValueError("Only modes {} are supported for 4D inputs".format(permitted_4_channel_modes))

        if mode is None and npimg.dtype == np.uint8:
            mode = 'RGBA'
    else:
        permitted_3_channel_modes = ['RGB', 'YCbCr', 'HSV']
        if mode is not None and mode not in permitted_3_channel_modes:
            raise ValueError("Only modes {} are supported for 3D inputs".format(permitted_3_channel_modes))
        if mode is None and npimg.dtype == np.uint8:
            mode = 'RGB'

    if mode is None:
        raise TypeError('Input type {} is not supported'.format(npimg.dtype))

    return Image.fromarray(npimg, mode=mode)


297
def normalize(tensor: Tensor, mean: List[float], std: List[float], inplace: bool = False) -> Tensor:
298
    """Normalize a tensor image with mean and standard deviation.
299
    This transform does not support PIL Image.
300

301
    .. note::
surgan12's avatar
surgan12 committed
302
        This transform acts out of place by default, i.e., it does not mutates the input tensor.
303

304
    See :class:`~torchvision.transforms.Normalize` for more details.
305
306

    Args:
307
        tensor (Tensor): Tensor image of size (C, H, W) or (B, C, H, W) to be normalized.
308
        mean (sequence): Sequence of means for each channel.
309
        std (sequence): Sequence of standard deviations for each channel.
310
        inplace(bool,optional): Bool to make this operation inplace.
311
312
313
314

    Returns:
        Tensor: Normalized Tensor image.
    """
315
316
    if not isinstance(tensor, torch.Tensor):
        raise TypeError('Input tensor should be a torch tensor. Got {}.'.format(type(tensor)))
317

318
319
    if tensor.ndim < 3:
        raise ValueError('Expected tensor to be a tensor image of size (..., C, H, W). Got tensor.size() = '
320
                         '{}.'.format(tensor.size()))
321

surgan12's avatar
surgan12 committed
322
323
324
    if not inplace:
        tensor = tensor.clone()

325
326
327
    dtype = tensor.dtype
    mean = torch.as_tensor(mean, dtype=dtype, device=tensor.device)
    std = torch.as_tensor(std, dtype=dtype, device=tensor.device)
328
329
    if (std == 0).any():
        raise ValueError('std evaluated to zero after conversion to {}, leading to division by zero.'.format(dtype))
330
    if mean.ndim == 1:
331
        mean = mean.view(-1, 1, 1)
332
    if std.ndim == 1:
333
        std = std.view(-1, 1, 1)
334
    tensor.sub_(mean).div_(std)
335
    return tensor
336
337


338
def resize(img: Tensor, size: List[int], interpolation: InterpolationMode = InterpolationMode.BILINEAR) -> Tensor:
vfdev's avatar
vfdev committed
339
    r"""Resize the input image to the given size.
340
    If the image is torch Tensor, it is expected
vfdev's avatar
vfdev committed
341
    to have [..., H, W] shape, where ... means an arbitrary number of leading dimensions
342
343

    Args:
vfdev's avatar
vfdev committed
344
        img (PIL Image or Tensor): Image to be resized.
345
346
        size (sequence or int): Desired output size. If size is a sequence like
            (h, w), the output size will be matched to this. If size is an int,
Vitaliy Chiley's avatar
Vitaliy Chiley committed
347
            the smaller edge of the image will be matched to this number maintaining
348
            the aspect ratio. i.e, if height > width, then image will be rescaled to
vfdev's avatar
vfdev committed
349
            :math:`\left(\text{size} \times \frac{\text{height}}{\text{width}}, \text{size}\right)`.
350
            In torchscript mode size as single int is not supported, use a sequence of length 1: ``[size, ]``.
351
352
353
354
        interpolation (InterpolationMode): Desired interpolation enum defined by
            :class:`torchvision.transforms.InterpolationMode`.
            Default is ``InterpolationMode.BILINEAR``. If input is Tensor, only ``InterpolationMode.NEAREST``,
            ``InterpolationMode.BILINEAR`` and ``InterpolationMode.BICUBIC`` are supported.
355
            For backward compatibility integer values (e.g. ``PIL.Image.NEAREST``) are still acceptable.
356
357

    Returns:
vfdev's avatar
vfdev committed
358
        PIL Image or Tensor: Resized image.
359
    """
360
361
362
    # Backward compatibility with integer value
    if isinstance(interpolation, int):
        warnings.warn(
363
364
            "Argument interpolation should be of type InterpolationMode instead of int. "
            "Please, use InterpolationMode enum."
365
366
367
        )
        interpolation = _interpolation_modes_from_int(interpolation)

368
369
    if not isinstance(interpolation, InterpolationMode):
        raise TypeError("Argument interpolation should be a InterpolationMode")
370

vfdev's avatar
vfdev committed
371
    if not isinstance(img, torch.Tensor):
372
373
        pil_interpolation = pil_modes_mapping[interpolation]
        return F_pil.resize(img, size=size, interpolation=pil_interpolation)
vfdev's avatar
vfdev committed
374

375
    return F_t.resize(img, size=size, interpolation=interpolation.value)
376
377
378
379
380
381
382
383


def scale(*args, **kwargs):
    warnings.warn("The use of the transforms.Scale transform is deprecated, " +
                  "please use transforms.Resize instead.")
    return resize(*args, **kwargs)


384
385
def pad(img: Tensor, padding: List[int], fill: int = 0, padding_mode: str = "constant") -> Tensor:
    r"""Pad the given image on all sides with the given "pad" value.
386
    If the image is torch Tensor, it is expected
387
    to have [..., H, W] shape, where ... means an arbitrary number of leading dimensions
388
389

    Args:
390
        img (PIL Image or Tensor): Image to be padded.
391
392
393
        padding (int or sequence): Padding on each border. If a single int is provided this
            is used to pad all borders. If sequence of length 2 is provided this is the padding
            on left/right and top/bottom respectively. If a sequence of length 4 is provided
394
            this is the padding for the left, top, right and bottom borders respectively.
395
396
397
398
399
400
            In torchscript mode padding as single int is not supported, use a sequence of length 1: ``[padding, ]``.
        fill (number or str or tuple): Pixel fill value for constant fill. Default is 0.
            If a tuple of length 3, it is used to fill R, G, B channels respectively.
            This value is only used when the padding_mode is constant.
            Only number is supported for torch Tensor.
            Only int or str or tuple value is supported for PIL Image.
401
        padding_mode: Type of padding. Should be: constant, edge, reflect or symmetric. Default is constant.
402
403
404
405
406
407
408
409
410
411
412
413
414
415

            - constant: pads with a constant value, this value is specified with fill

            - edge: pads with the last value on the edge of the image

            - reflect: pads with reflection of image (without repeating the last value on the edge)

                       padding [1, 2, 3, 4] with 2 elements on both sides in reflect mode
                       will result in [3, 2, 1, 2, 3, 4, 3, 2]

            - symmetric: pads with reflection of image (repeating the last value on the edge)

                         padding [1, 2, 3, 4] with 2 elements on both sides in symmetric mode
                         will result in [2, 1, 1, 2, 3, 4, 4, 3]
416
417

    Returns:
418
        PIL Image or Tensor: Padded image.
419
    """
420
421
    if not isinstance(img, torch.Tensor):
        return F_pil.pad(img, padding=padding, fill=fill, padding_mode=padding_mode)
422

423
    return F_t.pad(img, padding=padding, fill=fill, padding_mode=padding_mode)
424
425


vfdev's avatar
vfdev committed
426
427
def crop(img: Tensor, top: int, left: int, height: int, width: int) -> Tensor:
    """Crop the given image at specified location and output size.
428
429
    If the image is torch Tensor, it is expected
    to have [..., H, W] shape, where ... means an arbitrary number of leading dimensions
430

431
    Args:
vfdev's avatar
vfdev committed
432
        img (PIL Image or Tensor): Image to be cropped. (0,0) denotes the top left corner of the image.
433
434
435
436
        top (int): Vertical component of the top left corner of the crop box.
        left (int): Horizontal component of the top left corner of the crop box.
        height (int): Height of the crop box.
        width (int): Width of the crop box.
437

438
    Returns:
vfdev's avatar
vfdev committed
439
        PIL Image or Tensor: Cropped image.
440
441
    """

vfdev's avatar
vfdev committed
442
443
    if not isinstance(img, torch.Tensor):
        return F_pil.crop(img, top, left, height, width)
444

vfdev's avatar
vfdev committed
445
    return F_t.crop(img, top, left, height, width)
446

vfdev's avatar
vfdev committed
447
448
449

def center_crop(img: Tensor, output_size: List[int]) -> Tensor:
    """Crops the given image at the center.
450
    If the image is torch Tensor, it is expected
vfdev's avatar
vfdev committed
451
    to have [..., H, W] shape, where ... means an arbitrary number of leading dimensions
452

453
    Args:
vfdev's avatar
vfdev committed
454
        img (PIL Image or Tensor): Image to be cropped.
455
        output_size (sequence or int): (height, width) of the crop box. If int or sequence with single int,
vfdev's avatar
vfdev committed
456
457
            it is used for both directions.

458
    Returns:
vfdev's avatar
vfdev committed
459
        PIL Image or Tensor: Cropped image.
460
    """
461
462
    if isinstance(output_size, numbers.Number):
        output_size = (int(output_size), int(output_size))
vfdev's avatar
vfdev committed
463
464
465
466
    elif isinstance(output_size, (tuple, list)) and len(output_size) == 1:
        output_size = (output_size[0], output_size[0])

    image_width, image_height = _get_image_size(img)
467
    crop_height, crop_width = output_size
vfdev's avatar
vfdev committed
468

469
470
    crop_top = int(round((image_height - crop_height) / 2.))
    crop_left = int(round((image_width - crop_width) / 2.))
471
    return crop(img, crop_top, crop_left, crop_height, crop_width)
472
473


474
def resized_crop(
475
        img: Tensor, top: int, left: int, height: int, width: int, size: List[int],
476
        interpolation: InterpolationMode = InterpolationMode.BILINEAR
477
478
) -> Tensor:
    """Crop the given image and resize it to desired size.
479
    If the image is torch Tensor, it is expected
480
    to have [..., H, W] shape, where ... means an arbitrary number of leading dimensions
481

482
    Notably used in :class:`~torchvision.transforms.RandomResizedCrop`.
483
484

    Args:
485
        img (PIL Image or Tensor): Image to be cropped. (0,0) denotes the top left corner of the image.
486
487
488
489
        top (int): Vertical component of the top left corner of the crop box.
        left (int): Horizontal component of the top left corner of the crop box.
        height (int): Height of the crop box.
        width (int): Width of the crop box.
490
        size (sequence or int): Desired output size. Same semantics as ``resize``.
491
492
493
494
        interpolation (InterpolationMode): Desired interpolation enum defined by
            :class:`torchvision.transforms.InterpolationMode`.
            Default is ``InterpolationMode.BILINEAR``. If input is Tensor, only ``InterpolationMode.NEAREST``,
            ``InterpolationMode.BILINEAR`` and ``InterpolationMode.BICUBIC`` are supported.
495
496
            For backward compatibility integer values (e.g. ``PIL.Image.NEAREST``) are still acceptable.

497
    Returns:
498
        PIL Image or Tensor: Cropped image.
499
    """
500
    img = crop(img, top, left, height, width)
501
502
503
504
    img = resize(img, size, interpolation)
    return img


505
def hflip(img: Tensor) -> Tensor:
506
    """Horizontally flip the given image.
507
508

    Args:
vfdev's avatar
vfdev committed
509
        img (PIL Image or Tensor): Image to be flipped. If img
510
            is a Tensor, it is expected to be in [..., H, W] format,
511
            where ... means it can have an arbitrary number of leading
512
            dimensions.
513
514

    Returns:
vfdev's avatar
vfdev committed
515
        PIL Image or Tensor:  Horizontally flipped image.
516
    """
517
518
    if not isinstance(img, torch.Tensor):
        return F_pil.hflip(img)
519

520
    return F_t.hflip(img)
521
522


523
524
525
def _get_perspective_coeffs(
        startpoints: List[List[int]], endpoints: List[List[int]]
) -> List[float]:
526
527
    """Helper function to get the coefficients (a, b, c, d, e, f, g, h) for the perspective transforms.

Vitaliy Chiley's avatar
Vitaliy Chiley committed
528
    In Perspective Transform each pixel (x, y) in the original image gets transformed as,
529
530
531
     (x, y) -> ( (ax + by + c) / (gx + hy + 1), (dx + ey + f) / (gx + hy + 1) )

    Args:
532
533
534
535
536
        startpoints (list of list of ints): List containing four lists of two integers corresponding to four corners
            ``[top-left, top-right, bottom-right, bottom-left]`` of the original image.
        endpoints (list of list of ints): List containing four lists of two integers corresponding to four corners
            ``[top-left, top-right, bottom-right, bottom-left]`` of the transformed image.

537
538
539
    Returns:
        octuple (a, b, c, d, e, f, g, h) for transforming each pixel.
    """
540
541
542
543
544
    a_matrix = torch.zeros(2 * len(startpoints), 8, dtype=torch.float)

    for i, (p1, p2) in enumerate(zip(endpoints, startpoints)):
        a_matrix[2 * i, :] = torch.tensor([p1[0], p1[1], 1, 0, 0, 0, -p2[0] * p1[0], -p2[0] * p1[1]])
        a_matrix[2 * i + 1, :] = torch.tensor([0, 0, 0, p1[0], p1[1], 1, -p2[1] * p1[0], -p2[1] * p1[1]])
545

546
547
    b_matrix = torch.tensor(startpoints, dtype=torch.float).view(8)
    res = torch.lstsq(b_matrix, a_matrix)[0]
548

549
550
    output: List[float] = res.squeeze(1).tolist()
    return output
551
552


553
554
555
556
def perspective(
        img: Tensor,
        startpoints: List[List[int]],
        endpoints: List[List[int]],
557
        interpolation: InterpolationMode = InterpolationMode.BILINEAR,
558
        fill: Optional[List[float]] = None
559
560
) -> Tensor:
    """Perform perspective transform of the given image.
561
    If the image is torch Tensor, it is expected
562
    to have [..., H, W] shape, where ... means an arbitrary number of leading dimensions.
563
564

    Args:
565
566
567
568
569
        img (PIL Image or Tensor): Image to be transformed.
        startpoints (list of list of ints): List containing four lists of two integers corresponding to four corners
            ``[top-left, top-right, bottom-right, bottom-left]`` of the original image.
        endpoints (list of list of ints): List containing four lists of two integers corresponding to four corners
            ``[top-left, top-right, bottom-right, bottom-left]`` of the transformed image.
570
571
572
        interpolation (InterpolationMode): Desired interpolation enum defined by
            :class:`torchvision.transforms.InterpolationMode`. Default is ``InterpolationMode.BILINEAR``.
            If input is Tensor, only ``InterpolationMode.NEAREST``, ``InterpolationMode.BILINEAR`` are supported.
573
            For backward compatibility integer values (e.g. ``PIL.Image.NEAREST``) are still acceptable.
574
575
576
577
        fill (sequence or number, optional): Pixel fill value for the area outside the transformed
            image. If given a number, the value is used for all bands respectively.
            In torchscript mode single int/float value is not supported, please use a sequence
            of length 1: ``[value, ]``.
578
            If input is PIL Image, the options is only available for ``Pillow>=5.0.0``.
579

580
    Returns:
581
        PIL Image or Tensor: transformed Image.
582
    """
583

584
    coeffs = _get_perspective_coeffs(startpoints, endpoints)
585

586
587
588
    # Backward compatibility with integer value
    if isinstance(interpolation, int):
        warnings.warn(
589
590
            "Argument interpolation should be of type InterpolationMode instead of int. "
            "Please, use InterpolationMode enum."
591
592
593
        )
        interpolation = _interpolation_modes_from_int(interpolation)

594
595
    if not isinstance(interpolation, InterpolationMode):
        raise TypeError("Argument interpolation should be a InterpolationMode")
596

597
    if not isinstance(img, torch.Tensor):
598
599
        pil_interpolation = pil_modes_mapping[interpolation]
        return F_pil.perspective(img, coeffs, interpolation=pil_interpolation, fill=fill)
600

601
    return F_t.perspective(img, coeffs, interpolation=interpolation.value, fill=fill)
602
603


604
def vflip(img: Tensor) -> Tensor:
605
    """Vertically flip the given image.
606
607

    Args:
vfdev's avatar
vfdev committed
608
        img (PIL Image or Tensor): Image to be flipped. If img
609
            is a Tensor, it is expected to be in [..., H, W] format,
610
            where ... means it can have an arbitrary number of leading
611
            dimensions.
612
613

    Returns:
614
        PIL Image or Tensor:  Vertically flipped image.
615
    """
616
617
    if not isinstance(img, torch.Tensor):
        return F_pil.vflip(img)
618

619
    return F_t.vflip(img)
620
621


vfdev's avatar
vfdev committed
622
623
def five_crop(img: Tensor, size: List[int]) -> Tuple[Tensor, Tensor, Tensor, Tensor, Tensor]:
    """Crop the given image into four corners and the central crop.
624
    If the image is torch Tensor, it is expected
vfdev's avatar
vfdev committed
625
    to have [..., H, W] shape, where ... means an arbitrary number of leading dimensions
626
627
628
629
630
631

    .. Note::
        This transform returns a tuple of images and there may be a
        mismatch in the number of inputs and targets your ``Dataset`` returns.

    Args:
vfdev's avatar
vfdev committed
632
633
634
        img (PIL Image or Tensor): Image to be cropped.
        size (sequence or int): Desired output size of the crop. If size is an
            int instead of sequence like (h, w), a square crop (size, size) is
635
            made. If provided a sequence of length 1, it will be interpreted as (size[0], size[0]).
636

637
    Returns:
638
639
       tuple: tuple (tl, tr, bl, br, center)
                Corresponding top left, top right, bottom left, bottom right and center crop.
640
641
642
    """
    if isinstance(size, numbers.Number):
        size = (int(size), int(size))
vfdev's avatar
vfdev committed
643
644
    elif isinstance(size, (tuple, list)) and len(size) == 1:
        size = (size[0], size[0])
645

vfdev's avatar
vfdev committed
646
647
648
649
    if len(size) != 2:
        raise ValueError("Please provide only two dimensions (h, w) for size.")

    image_width, image_height = _get_image_size(img)
650
651
652
653
654
    crop_height, crop_width = size
    if crop_width > image_width or crop_height > image_height:
        msg = "Requested crop size {} is bigger than input size {}"
        raise ValueError(msg.format(size, (image_height, image_width)))

vfdev's avatar
vfdev committed
655
656
657
658
659
660
661
662
    tl = crop(img, 0, 0, crop_height, crop_width)
    tr = crop(img, 0, image_width - crop_width, crop_height, crop_width)
    bl = crop(img, image_height - crop_height, 0, crop_height, crop_width)
    br = crop(img, image_height - crop_height, image_width - crop_width, crop_height, crop_width)

    center = center_crop(img, [crop_height, crop_width])

    return tl, tr, bl, br, center
663
664


vfdev's avatar
vfdev committed
665
666
667
def ten_crop(img: Tensor, size: List[int], vertical_flip: bool = False) -> List[Tensor]:
    """Generate ten cropped images from the given image.
    Crop the given image into four corners and the central crop plus the
668
    flipped version of these (horizontal flipping is used by default).
669
    If the image is torch Tensor, it is expected
vfdev's avatar
vfdev committed
670
    to have [..., H, W] shape, where ... means an arbitrary number of leading dimensions
671
672
673
674
675

    .. Note::
        This transform returns a tuple of images and there may be a
        mismatch in the number of inputs and targets your ``Dataset`` returns.

676
    Args:
vfdev's avatar
vfdev committed
677
        img (PIL Image or Tensor): Image to be cropped.
678
        size (sequence or int): Desired output size of the crop. If size is an
679
            int instead of sequence like (h, w), a square crop (size, size) is
680
            made. If provided a sequence of length 1, it will be interpreted as (size[0], size[0]).
681
        vertical_flip (bool): Use vertical flipping instead of horizontal
682
683

    Returns:
684
685
686
        tuple: tuple (tl, tr, bl, br, center, tl_flip, tr_flip, bl_flip, br_flip, center_flip)
            Corresponding top left, top right, bottom left, bottom right and
            center crop and same for the flipped image.
687
688
689
    """
    if isinstance(size, numbers.Number):
        size = (int(size), int(size))
vfdev's avatar
vfdev committed
690
691
692
693
694
    elif isinstance(size, (tuple, list)) and len(size) == 1:
        size = (size[0], size[0])

    if len(size) != 2:
        raise ValueError("Please provide only two dimensions (h, w) for size.")
695
696
697
698
699
700
701
702
703
704
705
706

    first_five = five_crop(img, size)

    if vertical_flip:
        img = vflip(img)
    else:
        img = hflip(img)

    second_five = five_crop(img, size)
    return first_five + second_five


707
def adjust_brightness(img: Tensor, brightness_factor: float) -> Tensor:
708
    """Adjust brightness of an image.
709
710

    Args:
vfdev's avatar
vfdev committed
711
        img (PIL Image or Tensor): Image to be adjusted.
712
713
        If img is a Tensor, it is expected to be in [..., 1 or 3, H, W] format,
        where ... means it can have an arbitrary number of leading dimensions.
714
715
716
717
718
        brightness_factor (float):  How much to adjust the brightness. Can be
            any non negative number. 0 gives a black image, 1 gives the
            original image while 2 increases the brightness by a factor of 2.

    Returns:
vfdev's avatar
vfdev committed
719
        PIL Image or Tensor: Brightness adjusted image.
720
    """
721
722
    if not isinstance(img, torch.Tensor):
        return F_pil.adjust_brightness(img, brightness_factor)
723

724
    return F_t.adjust_brightness(img, brightness_factor)
725
726


727
def adjust_contrast(img: Tensor, contrast_factor: float) -> Tensor:
728
    """Adjust contrast of an image.
729
730

    Args:
vfdev's avatar
vfdev committed
731
        img (PIL Image or Tensor): Image to be adjusted.
732
733
734
735
736
        contrast_factor (float): How much to adjust the contrast. Can be any
            non negative number. 0 gives a solid gray image, 1 gives the
            original image while 2 increases the contrast by a factor of 2.

    Returns:
vfdev's avatar
vfdev committed
737
        PIL Image or Tensor: Contrast adjusted image.
738
    """
739
740
    if not isinstance(img, torch.Tensor):
        return F_pil.adjust_contrast(img, contrast_factor)
741

742
    return F_t.adjust_contrast(img, contrast_factor)
743
744


745
def adjust_saturation(img: Tensor, saturation_factor: float) -> Tensor:
746
747
748
    """Adjust color saturation of an image.

    Args:
vfdev's avatar
vfdev committed
749
        img (PIL Image or Tensor): Image to be adjusted.
750
751
752
753
754
        saturation_factor (float):  How much to adjust the saturation. 0 will
            give a black and white image, 1 will give the original image while
            2 will enhance the saturation by a factor of 2.

    Returns:
vfdev's avatar
vfdev committed
755
        PIL Image or Tensor: Saturation adjusted image.
756
    """
757
758
    if not isinstance(img, torch.Tensor):
        return F_pil.adjust_saturation(img, saturation_factor)
759

760
    return F_t.adjust_saturation(img, saturation_factor)
761
762


763
def adjust_hue(img: Tensor, hue_factor: float) -> Tensor:
764
765
766
767
768
769
770
771
772
    """Adjust hue of an image.

    The image hue is adjusted by converting the image to HSV and
    cyclically shifting the intensities in the hue channel (H).
    The image is then converted back to original image mode.

    `hue_factor` is the amount of shift in H channel and must be in the
    interval `[-0.5, 0.5]`.

773
774
775
    See `Hue`_ for more details.

    .. _Hue: https://en.wikipedia.org/wiki/Hue
776
777

    Args:
778
        img (PIL Image or Tensor): Image to be adjusted.
779
780
781
782
783
784
785
        hue_factor (float):  How much to shift the hue channel. Should be in
            [-0.5, 0.5]. 0.5 and -0.5 give complete reversal of hue channel in
            HSV space in positive and negative direction respectively.
            0 means no shift. Therefore, both -0.5 and 0.5 will give an image
            with complementary colors while 0 gives the original image.

    Returns:
786
        PIL Image or Tensor: Hue adjusted image.
787
    """
788
789
    if not isinstance(img, torch.Tensor):
        return F_pil.adjust_hue(img, hue_factor)
790

791
    return F_t.adjust_hue(img, hue_factor)
792
793


794
def adjust_gamma(img: Tensor, gamma: float, gain: float = 1) -> Tensor:
795
    r"""Perform gamma correction on an image.
796
797
798
799

    Also known as Power Law Transform. Intensities in RGB mode are adjusted
    based on the following equation:

800
801
802
803
    .. math::
        I_{\text{out}} = 255 \times \text{gain} \times \left(\frac{I_{\text{in}}}{255}\right)^{\gamma}

    See `Gamma Correction`_ for more details.
804

805
    .. _Gamma Correction: https://en.wikipedia.org/wiki/Gamma_correction
806
807

    Args:
808
        img (PIL Image or Tensor): PIL Image to be adjusted.
809
810
        If img is a Tensor, it is expected to be in [..., 1 or 3, H, W] format,
        where ... means it can have an arbitrary number of leading dimensions.
811
812
813
        gamma (float): Non negative real number, same as :math:`\gamma` in the equation.
            gamma larger than 1 make the shadows darker,
            while gamma smaller than 1 make dark regions lighter.
814
        gain (float): The constant multiplier.
815
816
    Returns:
        PIL Image or Tensor: Gamma correction adjusted image.
817
    """
818
819
    if not isinstance(img, torch.Tensor):
        return F_pil.adjust_gamma(img, gamma, gain)
820

821
    return F_t.adjust_gamma(img, gamma, gain)
822
823


vfdev's avatar
vfdev committed
824
def _get_inverse_affine_matrix(
vfdev's avatar
vfdev committed
825
        center: List[float], angle: float, translate: List[float], scale: float, shear: List[float]
vfdev's avatar
vfdev committed
826
) -> List[float]:
827
828
829
830
831
832
833
    # Helper method to compute inverse matrix for affine transformation

    # As it is explained in PIL.Image.rotate
    # We need compute INVERSE of affine transformation matrix: M = T * C * RSS * C^-1
    # where T is translation matrix: [1, 0, tx | 0, 1, ty | 0, 0, 1]
    #       C is translation matrix to keep center: [1, 0, cx | 0, 1, cy | 0, 0, 1]
    #       RSS is rotation with scale and shear matrix
834
835
836
837
838
839
840
841
842
843
    #       RSS(a, s, (sx, sy)) =
    #       = R(a) * S(s) * SHy(sy) * SHx(sx)
    #       = [ s*cos(a - sy)/cos(sy), s*(-cos(a - sy)*tan(x)/cos(y) - sin(a)), 0 ]
    #         [ s*sin(a + sy)/cos(sy), s*(-sin(a - sy)*tan(x)/cos(y) + cos(a)), 0 ]
    #         [ 0                    , 0                                      , 1 ]
    #
    # where R is a rotation matrix, S is a scaling matrix, and SHx and SHy are the shears:
    # SHx(s) = [1, -tan(s)] and SHy(s) = [1      , 0]
    #          [0, 1      ]              [-tan(s), 1]
    #
844
845
    # Thus, the inverse is M^-1 = C * RSS^-1 * C^-1 * T^-1

846
847
848
849
850
851
852
    rot = math.radians(angle)
    sx, sy = [math.radians(s) for s in shear]

    cx, cy = center
    tx, ty = translate

    # RSS without scaling
vfdev's avatar
vfdev committed
853
854
855
856
    a = math.cos(rot - sy) / math.cos(sy)
    b = -math.cos(rot - sy) * math.tan(sx) / math.cos(sy) - math.sin(rot)
    c = math.sin(rot - sy) / math.cos(sy)
    d = -math.sin(rot - sy) * math.tan(sx) / math.cos(sy) + math.cos(rot)
857
858

    # Inverted rotation matrix with scale and shear
859
    # det([[a, b], [c, d]]) == 1, since det(rotation) = 1 and det(shear) = 1
vfdev's avatar
vfdev committed
860
861
    matrix = [d, -b, 0.0, -c, a, 0.0]
    matrix = [x / scale for x in matrix]
862
863

    # Apply inverse of translation and of center translation: RSS^-1 * C^-1 * T^-1
vfdev's avatar
vfdev committed
864
865
    matrix[2] += matrix[0] * (-cx - tx) + matrix[1] * (-cy - ty)
    matrix[5] += matrix[3] * (-cx - tx) + matrix[4] * (-cy - ty)
866
867

    # Apply center translation: C * RSS^-1 * C^-1 * T^-1
vfdev's avatar
vfdev committed
868
869
    matrix[2] += cx
    matrix[5] += cy
870

vfdev's avatar
vfdev committed
871
    return matrix
872

vfdev's avatar
vfdev committed
873

vfdev's avatar
vfdev committed
874
def rotate(
875
        img: Tensor, angle: float, interpolation: InterpolationMode = InterpolationMode.NEAREST,
876
        expand: bool = False, center: Optional[List[int]] = None,
877
        fill: Optional[List[float]] = None, resample: Optional[int] = None
vfdev's avatar
vfdev committed
878
879
) -> Tensor:
    """Rotate the image by angle.
880
    If the image is torch Tensor, it is expected
vfdev's avatar
vfdev committed
881
882
883
884
    to have [..., H, W] shape, where ... means an arbitrary number of leading dimensions.

    Args:
        img (PIL Image or Tensor): image to be rotated.
885
        angle (number): rotation angle value in degrees, counter-clockwise.
886
887
888
        interpolation (InterpolationMode): Desired interpolation enum defined by
            :class:`torchvision.transforms.InterpolationMode`. Default is ``InterpolationMode.NEAREST``.
            If input is Tensor, only ``InterpolationMode.NEAREST``, ``InterpolationMode.BILINEAR`` are supported.
889
            For backward compatibility integer values (e.g. ``PIL.Image.NEAREST``) are still acceptable.
vfdev's avatar
vfdev committed
890
891
892
893
        expand (bool, optional): Optional expansion flag.
            If true, expands the output image to make it large enough to hold the entire rotated image.
            If false or omitted, make the output image the same size as the input image.
            Note that the expand flag assumes rotation around the center and no translation.
894
        center (sequence, optional): Optional center of rotation. Origin is the upper left corner.
vfdev's avatar
vfdev committed
895
            Default is the center of the image.
896
897
898
899
        fill (sequence or number, optional): Pixel fill value for the area outside the transformed
            image. If given a number, the value is used for all bands respectively.
            In torchscript mode single int/float value is not supported, please use a sequence
            of length 1: ``[value, ]``.
900
            If input is PIL Image, the options is only available for ``Pillow>=5.2.0``.
vfdev's avatar
vfdev committed
901
902
903
904
905
906
907

    Returns:
        PIL Image or Tensor: Rotated image.

    .. _filters: https://pillow.readthedocs.io/en/latest/handbook/concepts.html#filters

    """
908
909
910
911
912
913
914
915
916
    if resample is not None:
        warnings.warn(
            "Argument resample is deprecated and will be removed since v0.10.0. Please, use interpolation instead"
        )
        interpolation = _interpolation_modes_from_int(resample)

    # Backward compatibility with integer value
    if isinstance(interpolation, int):
        warnings.warn(
917
918
            "Argument interpolation should be of type InterpolationMode instead of int. "
            "Please, use InterpolationMode enum."
919
920
921
        )
        interpolation = _interpolation_modes_from_int(interpolation)

vfdev's avatar
vfdev committed
922
923
924
925
926
927
    if not isinstance(angle, (int, float)):
        raise TypeError("Argument angle should be int or float")

    if center is not None and not isinstance(center, (list, tuple)):
        raise TypeError("Argument center should be a sequence")

928
929
    if not isinstance(interpolation, InterpolationMode):
        raise TypeError("Argument interpolation should be a InterpolationMode")
930

vfdev's avatar
vfdev committed
931
    if not isinstance(img, torch.Tensor):
932
933
        pil_interpolation = pil_modes_mapping[interpolation]
        return F_pil.rotate(img, angle=angle, interpolation=pil_interpolation, expand=expand, center=center, fill=fill)
vfdev's avatar
vfdev committed
934
935
936
937

    center_f = [0.0, 0.0]
    if center is not None:
        img_size = _get_image_size(img)
938
939
940
        # 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, img_size)]

vfdev's avatar
vfdev committed
941
942
943
    # 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])
944
    return F_t.rotate(img, matrix=matrix, interpolation=interpolation.value, expand=expand, fill=fill)
vfdev's avatar
vfdev committed
945
946


vfdev's avatar
vfdev committed
947
948
def affine(
        img: Tensor, angle: float, translate: List[int], scale: float, shear: List[float],
949
950
        interpolation: InterpolationMode = InterpolationMode.NEAREST, fill: Optional[List[float]] = None,
        resample: Optional[int] = None, fillcolor: Optional[List[float]] = None
vfdev's avatar
vfdev committed
951
952
) -> Tensor:
    """Apply affine transformation on the image keeping image center invariant.
953
    If the image is torch Tensor, it is expected
vfdev's avatar
vfdev committed
954
    to have [..., H, W] shape, where ... means an arbitrary number of leading dimensions.
955
956

    Args:
vfdev's avatar
vfdev committed
957
        img (PIL Image or Tensor): image to transform.
958
959
        angle (number): rotation angle in degrees between -180 and 180, clockwise direction.
        translate (sequence of integers): horizontal and vertical translations (post-rotation translation)
960
        scale (float): overall scale
961
962
        shear (float or sequence): shear angle value in degrees between -180 to 180, clockwise direction.
            If a sequence is specified, the first value corresponds to a shear parallel to the x axis, while
vfdev's avatar
vfdev committed
963
            the second value corresponds to a shear parallel to the y axis.
964
965
966
        interpolation (InterpolationMode): Desired interpolation enum defined by
            :class:`torchvision.transforms.InterpolationMode`. Default is ``InterpolationMode.NEAREST``.
            If input is Tensor, only ``InterpolationMode.NEAREST``, ``InterpolationMode.BILINEAR`` are supported.
967
            For backward compatibility integer values (e.g. ``PIL.Image.NEAREST``) are still acceptable.
968
969
970
971
        fill (sequence or number, optional): Pixel fill value for the area outside the transformed
            image. If given a number, the value is used for all bands respectively.
            In torchscript mode single int/float value is not supported, please use a sequence
            of length 1: ``[value, ]``.
972
973
            If input is PIL Image, the options is only available for ``Pillow>=5.0.0``.
        fillcolor (sequence, int, float): deprecated argument and will be removed since v0.10.0.
974
975
976
            Please use `arg`:fill: instead.
        resample (int, optional): deprecated argument and will be removed since v0.10.0.
            Please use `arg`:interpolation: instead.
vfdev's avatar
vfdev committed
977
978
979

    Returns:
        PIL Image or Tensor: Transformed image.
980
    """
981
982
983
984
985
986
987
988
989
    if resample is not None:
        warnings.warn(
            "Argument resample is deprecated and will be removed since v0.10.0. Please, use interpolation instead"
        )
        interpolation = _interpolation_modes_from_int(resample)

    # Backward compatibility with integer value
    if isinstance(interpolation, int):
        warnings.warn(
990
991
            "Argument interpolation should be of type InterpolationMode instead of int. "
            "Please, use InterpolationMode enum."
992
993
994
995
996
997
998
999
1000
        )
        interpolation = _interpolation_modes_from_int(interpolation)

    if fillcolor is not None:
        warnings.warn(
            "Argument fillcolor is deprecated and will be removed since v0.10.0. Please, use fill instead"
        )
        fill = fillcolor

vfdev's avatar
vfdev committed
1001
1002
1003
1004
1005
1006
1007
1008
1009
1010
1011
1012
1013
1014
1015
    if not isinstance(angle, (int, float)):
        raise TypeError("Argument angle should be int or float")

    if not isinstance(translate, (list, tuple)):
        raise TypeError("Argument translate should be a sequence")

    if len(translate) != 2:
        raise ValueError("Argument translate should be a sequence of length 2")

    if scale <= 0.0:
        raise ValueError("Argument scale should be positive")

    if not isinstance(shear, (numbers.Number, (list, tuple))):
        raise TypeError("Shear should be either a single value or a sequence of two values")

1016
1017
    if not isinstance(interpolation, InterpolationMode):
        raise TypeError("Argument interpolation should be a InterpolationMode")
1018

vfdev's avatar
vfdev committed
1019
1020
1021
1022
1023
1024
1025
1026
1027
1028
1029
1030
1031
1032
1033
1034
1035
1036
1037
1038
1039
1040
1041
1042
1043
    if isinstance(angle, int):
        angle = float(angle)

    if isinstance(translate, tuple):
        translate = list(translate)

    if isinstance(shear, numbers.Number):
        shear = [shear, 0.0]

    if isinstance(shear, tuple):
        shear = list(shear)

    if len(shear) == 1:
        shear = [shear[0], shear[0]]

    if len(shear) != 2:
        raise ValueError("Shear should be a sequence containing two values. Got {}".format(shear))

    img_size = _get_image_size(img)
    if not isinstance(img, torch.Tensor):
        # center = (img_size[0] * 0.5 + 0.5, img_size[1] * 0.5 + 0.5)
        # it is visually better to estimate the center without 0.5 offset
        # otherwise image rotated by 90 degrees is shifted vs output image of torch.rot90 or F_t.affine
        center = [img_size[0] * 0.5, img_size[1] * 0.5]
        matrix = _get_inverse_affine_matrix(center, angle, translate, scale, shear)
1044
1045
        pil_interpolation = pil_modes_mapping[interpolation]
        return F_pil.affine(img, matrix=matrix, interpolation=pil_interpolation, fill=fill)
1046

1047
1048
    translate_f = [1.0 * t for t in translate]
    matrix = _get_inverse_affine_matrix([0.0, 0.0], angle, translate_f, scale, shear)
1049
    return F_t.affine(img, matrix=matrix, interpolation=interpolation.value, fill=fill)
1050
1051


1052
@torch.jit.unused
1053
def to_grayscale(img, num_output_channels=1):
1054
    """Convert PIL image of any mode (RGB, HSV, LAB, etc) to grayscale version of image.
1055
    This transform does not support torch Tensor.
1056
1057

    Args:
1058
        img (PIL Image): PIL Image to be converted to grayscale.
1059
        num_output_channels (int): number of channels of the output image. Value can be 1 or 3. Default is 1.
1060
1061

    Returns:
1062
1063
1064
1065
        PIL Image: Grayscale version of the image.
            if num_output_channels = 1 : returned image is single channel

            if num_output_channels = 3 : returned image is 3 channel with r = g = b
1066
    """
1067
1068
    if isinstance(img, Image.Image):
        return F_pil.to_grayscale(img, num_output_channels)
1069

1070
1071
1072
1073
1074
    raise TypeError("Input should be PIL Image")


def rgb_to_grayscale(img: Tensor, num_output_channels: int = 1) -> Tensor:
    """Convert RGB image to grayscale version of image.
1075
1076
    If the image is torch Tensor, it is expected
    to have [..., 3, H, W] shape, where ... means an arbitrary number of leading dimensions
1077
1078
1079
1080
1081
1082
1083
1084
1085
1086
1087
1088
1089
1090
1091
1092
1093
1094
1095

    Note:
        Please, note that this method supports only RGB images as input. For inputs in other color spaces,
        please, consider using meth:`~torchvision.transforms.functional.to_grayscale` with PIL Image.

    Args:
        img (PIL Image or Tensor): RGB Image to be converted to grayscale.
        num_output_channels (int): number of channels of the output image. Value can be 1 or 3. Default, 1.

    Returns:
        PIL Image or Tensor: Grayscale version of the image.
            if num_output_channels = 1 : returned image is single channel

            if num_output_channels = 3 : returned image is 3 channel with r = g = b
    """
    if not isinstance(img, torch.Tensor):
        return F_pil.to_grayscale(img, num_output_channels)

    return F_t.rgb_to_grayscale(img, num_output_channels)
1096
1097


1098
def erase(img: Tensor, i: int, j: int, h: int, w: int, v: Tensor, inplace: bool = False) -> Tensor:
1099
    """ Erase the input Tensor Image with given value.
1100
    This transform does not support PIL Image.
1101
1102
1103
1104
1105
1106
1107
1108

    Args:
        img (Tensor Image): Tensor image of size (C, H, W) to be erased
        i (int): i in (i,j) i.e coordinates of the upper left corner.
        j (int): j in (i,j) i.e coordinates of the upper left corner.
        h (int): Height of the erased region.
        w (int): Width of the erased region.
        v: Erasing value.
Zhun Zhong's avatar
Zhun Zhong committed
1109
        inplace(bool, optional): For in-place operations. By default is set False.
1110
1111
1112
1113
1114
1115
1116

    Returns:
        Tensor Image: Erased image.
    """
    if not isinstance(img, torch.Tensor):
        raise TypeError('img should be Tensor Image. Got {}'.format(type(img)))

1117
1118
1119
    if not inplace:
        img = img.clone()

vfdev's avatar
vfdev committed
1120
    img[..., i:i + h, j:j + w] = v
1121
    return img
1122
1123
1124


def gaussian_blur(img: Tensor, kernel_size: List[int], sigma: Optional[List[float]] = None) -> Tensor:
1125
1126
1127
    """Performs Gaussian blurring on the image by given kernel.
    If the image is torch Tensor, it is expected
    to have [..., H, W] shape, where ... means an arbitrary number of leading dimensions.
1128
1129
1130
1131
1132

    Args:
        img (PIL Image or Tensor): Image to be blurred
        kernel_size (sequence of ints or int): Gaussian kernel size. Can be a sequence of integers
            like ``(kx, ky)`` or a single integer for square kernels.
1133
            In torchscript mode kernel_size as single int is not supported, use a sequence of length 1: ``[ksize, ]``.
1134
1135
1136
1137
1138
        sigma (sequence of floats or float, optional): Gaussian kernel standard deviation. Can be a
            sequence of floats like ``(sigma_x, sigma_y)`` or a single float to define the
            same sigma in both X/Y directions. If None, then it is computed using
            ``kernel_size`` as ``sigma = 0.3 * ((kernel_size - 1) * 0.5 - 1) + 0.8``.
            Default, None. In torchscript mode sigma as single float is
1139
            not supported, use a sequence of length 1: ``[sigma, ]``.
1140
1141
1142
1143
1144
1145
1146
1147
1148
1149
1150
1151
1152
1153
1154
1155
1156
1157
1158
1159
1160
1161
1162
1163
1164
1165
1166
1167
1168
1169
1170
1171
1172
1173
1174
1175
1176
1177
1178
1179
1180

    Returns:
        PIL Image or Tensor: Gaussian Blurred version of the image.
    """
    if not isinstance(kernel_size, (int, list, tuple)):
        raise TypeError('kernel_size should be int or a sequence of integers. Got {}'.format(type(kernel_size)))
    if isinstance(kernel_size, int):
        kernel_size = [kernel_size, kernel_size]
    if len(kernel_size) != 2:
        raise ValueError('If kernel_size is a sequence its length should be 2. Got {}'.format(len(kernel_size)))
    for ksize in kernel_size:
        if ksize % 2 == 0 or ksize < 0:
            raise ValueError('kernel_size should have odd and positive integers. Got {}'.format(kernel_size))

    if sigma is None:
        sigma = [ksize * 0.15 + 0.35 for ksize in kernel_size]

    if sigma is not None and not isinstance(sigma, (int, float, list, tuple)):
        raise TypeError('sigma should be either float or sequence of floats. Got {}'.format(type(sigma)))
    if isinstance(sigma, (int, float)):
        sigma = [float(sigma), float(sigma)]
    if isinstance(sigma, (list, tuple)) and len(sigma) == 1:
        sigma = [sigma[0], sigma[0]]
    if len(sigma) != 2:
        raise ValueError('If sigma is a sequence, its length should be 2. Got {}'.format(len(sigma)))
    for s in sigma:
        if s <= 0.:
            raise ValueError('sigma should have positive values. Got {}'.format(sigma))

    t_img = img
    if not isinstance(img, torch.Tensor):
        if not F_pil._is_pil_image(img):
            raise TypeError('img should be PIL Image or Tensor. Got {}'.format(type(img)))

        t_img = to_tensor(img)

    output = F_t.gaussian_blur(t_img, kernel_size, sigma)

    if not isinstance(img, torch.Tensor):
        output = to_pil_image(output)
    return output
1181
1182
1183


def invert(img: Tensor) -> Tensor:
1184
    """Invert the colors of an RGB/grayscale image.
1185
1186
1187

    Args:
        img (PIL Image or Tensor): Image to have its colors inverted.
1188
1189
1190
            If img is a Tensor, it is expected to be in [..., 1 or 3, H, W] format,
            where ... means it can have an arbitrary number of leading dimensions.
            If img is PIL Image, it is expected to be in mode "L" or "RGB".
1191
1192
1193
1194
1195
1196
1197
1198
1199
1200
1201

    Returns:
        PIL Image or Tensor: Color inverted image.
    """
    if not isinstance(img, torch.Tensor):
        return F_pil.invert(img)

    return F_t.invert(img)


def posterize(img: Tensor, bits: int) -> Tensor:
1202
    """Posterize an image by reducing the number of bits for each color channel.
1203
1204
1205
1206

    Args:
        img (PIL Image or Tensor): Image to have its colors posterized.
            If img is a Tensor, it should be of type torch.uint8 and
1207
1208
1209
            it is expected to be in [..., 1 or 3, H, W] format, where ... means
            it can have an arbitrary number of leading dimensions.
            If img is PIL Image, it is expected to be in mode "L" or "RGB".
1210
1211
1212
1213
1214
1215
1216
1217
1218
1219
1220
1221
1222
1223
        bits (int): The number of bits to keep for each channel (0-8).
    Returns:
        PIL Image or Tensor: Posterized image.
    """
    if not (0 <= bits <= 8):
        raise ValueError('The number if bits should be between 0 and 8. Got {}'.format(bits))

    if not isinstance(img, torch.Tensor):
        return F_pil.posterize(img, bits)

    return F_t.posterize(img, bits)


def solarize(img: Tensor, threshold: float) -> Tensor:
1224
    """Solarize an RGB/grayscale image by inverting all pixel values above a threshold.
1225
1226
1227

    Args:
        img (PIL Image or Tensor): Image to have its colors inverted.
1228
1229
1230
            If img is a Tensor, it is expected to be in [..., 1 or 3, H, W] format,
            where ... means it can have an arbitrary number of leading dimensions.
            If img is PIL Image, it is expected to be in mode "L" or "RGB".
1231
1232
1233
1234
1235
1236
1237
1238
1239
1240
1241
        threshold (float): All pixels equal or above this value are inverted.
    Returns:
        PIL Image or Tensor: Solarized image.
    """
    if not isinstance(img, torch.Tensor):
        return F_pil.solarize(img, threshold)

    return F_t.solarize(img, threshold)


def adjust_sharpness(img: Tensor, sharpness_factor: float) -> Tensor:
1242
    """Adjust the sharpness of an image.
1243
1244
1245

    Args:
        img (PIL Image or Tensor): Image to be adjusted.
1246
1247
        If img is a Tensor, it is expected to be in [..., 1 or 3, H, W] format,
        where ... means it can have an arbitrary number of leading dimensions.
1248
1249
1250
1251
1252
1253
1254
1255
1256
1257
1258
1259
1260
1261
        sharpness_factor (float):  How much to adjust the sharpness. Can be
            any non negative number. 0 gives a blurred image, 1 gives the
            original image while 2 increases the sharpness by a factor of 2.

    Returns:
        PIL Image or Tensor: Sharpness adjusted image.
    """
    if not isinstance(img, torch.Tensor):
        return F_pil.adjust_sharpness(img, sharpness_factor)

    return F_t.adjust_sharpness(img, sharpness_factor)


def autocontrast(img: Tensor) -> Tensor:
1262
    """Maximize contrast of an image by remapping its
1263
1264
1265
1266
1267
    pixels per channel so that the lowest becomes black and the lightest
    becomes white.

    Args:
        img (PIL Image or Tensor): Image on which autocontrast is applied.
1268
1269
1270
            If img is a Tensor, it is expected to be in [..., 1 or 3, H, W] format,
            where ... means it can have an arbitrary number of leading dimensions.
            If img is PIL Image, it is expected to be in mode "L" or "RGB".
1271
1272
1273
1274
1275
1276
1277
1278
1279
1280
1281

    Returns:
        PIL Image or Tensor: An image that was autocontrasted.
    """
    if not isinstance(img, torch.Tensor):
        return F_pil.autocontrast(img)

    return F_t.autocontrast(img)


def equalize(img: Tensor) -> Tensor:
1282
    """Equalize the histogram of an image by applying
1283
1284
1285
1286
1287
    a non-linear mapping to the input in order to create a uniform
    distribution of grayscale values in the output.

    Args:
        img (PIL Image or Tensor): Image on which equalize is applied.
1288
1289
1290
            If img is a Tensor, it is expected to be in [..., 1 or 3, H, W] format,
            where ... means it can have an arbitrary number of leading dimensions.
            If img is PIL Image, it is expected to be in mode "P", "L" or "RGB".
1291
1292
1293
1294
1295
1296
1297
1298

    Returns:
        PIL Image or Tensor: An image that was equalized.
    """
    if not isinstance(img, torch.Tensor):
        return F_pil.equalize(img)

    return F_t.equalize(img)