RGB<->Grayscale CV-CUDA Backend

test/test_transforms_v2.py

@@ -21,6 +21,7 @@
 import torchvision.transforms.v2 as transforms
 
 from common_utils import (
+    assert_close,
     assert_equal,
     cache,
     cpu_and_cuda,
@@ -41,7 +42,6 @@
 )
 
 from torch import nn
-from torch.testing import assert_close
 from torch.utils._pytree import tree_flatten, tree_map
 from torch.utils.data import DataLoader, default_collate
 from torchvision import tv_tensors
@@ -6409,7 +6409,17 @@ class TestRgbToGrayscale:
     def test_kernel_image(self, dtype, device):
         check_kernel(F.rgb_to_grayscale_image, make_image(dtype=dtype, device=device))
 
-    @pytest.mark.parametrize("make_input", [make_image_tensor, make_image_pil, make_image])
+    @pytest.mark.parametrize(
+        "make_input",
+        [
+            make_image_tensor,
+            make_image_pil,
+            make_image,
+            pytest.param(
+                make_image_cvcuda, marks=pytest.mark.skipif(not CVCUDA_AVAILABLE, reason="CVCUDA not available")
+            ),
+        ],
+    )
     def test_functional(self, make_input):
         check_functional(F.rgb_to_grayscale, make_input())
 
@@ -6419,23 +6429,53 @@ def test_functional(self, make_input):
             (F.rgb_to_grayscale_image, torch.Tensor),
             (F._color._rgb_to_grayscale_image_pil, PIL.Image.Image),
             (F.rgb_to_grayscale_image, tv_tensors.Image),
+            pytest.param(
+                F._color._rgb_to_grayscale_image_cvcuda,
+                None,
+                marks=pytest.mark.skipif(not CVCUDA_AVAILABLE, reason="CVCUDA not available"),
+            ),
         ],
     )
     def test_functional_signature(self, kernel, input_type):
+        if kernel is F._color._rgb_to_grayscale_image_cvcuda:
+            input_type = _import_cvcuda().Tensor
         check_functional_kernel_signature_match(F.rgb_to_grayscale, kernel=kernel, input_type=input_type)
 
     @pytest.mark.parametrize("transform", [transforms.Grayscale(), transforms.RandomGrayscale(p=1)])
-    @pytest.mark.parametrize("make_input", [make_image_tensor, make_image_pil, make_image])
+    @pytest.mark.parametrize(
+        "make_input",
+        [
+            make_image_tensor,
+            make_image_pil,
+            make_image,
+            pytest.param(
+                make_image_cvcuda, marks=pytest.mark.skipif(not CVCUDA_AVAILABLE, reason="CVCUDA not available")
+            ),
+        ],
+    )
     def test_transform(self, transform, make_input):
         check_transform(transform, make_input())
 
     @pytest.mark.parametrize("num_output_channels", [1, 3])
     @pytest.mark.parametrize("color_space", ["RGB", "GRAY"])
+    @pytest.mark.parametrize(
+        "make_input",
+        [
+            make_image,
+            pytest.param(
+                make_image_cvcuda, marks=pytest.mark.skipif(not CVCUDA_AVAILABLE, reason="CVCUDA not available")
+            ),
+        ],
+    )
     @pytest.mark.parametrize("fn", [F.rgb_to_grayscale, transform_cls_to_functional(transforms.Grayscale)])
-    def test_image_correctness(self, num_output_channels, color_space, fn):
-        image = make_image(dtype=torch.uint8, device="cpu", color_space=color_space)
+    def test_image_correctness(self, num_output_channels, color_space, make_input, fn):
+        image = make_input(dtype=torch.uint8, device="cpu", color_space=color_space)
 
         actual = fn(image, num_output_channels=num_output_channels)
+
+        if make_input is make_image_cvcuda:
+            image = F.cvcuda_to_tensor(image)[0].cpu()
+
         expected = F.to_image(F.rgb_to_grayscale(F.to_pil_image(image), num_output_channels=num_output_channels))
 
         assert_equal(actual, expected, rtol=0, atol=1)
@@ -6473,7 +6513,17 @@ class TestGrayscaleToRgb:
     def test_kernel_image(self, dtype, device):
         check_kernel(F.grayscale_to_rgb_image, make_image(dtype=dtype, device=device))
 
-    @pytest.mark.parametrize("make_input", [make_image_tensor, make_image_pil, make_image])
+    @pytest.mark.parametrize(
+        "make_input",
+        [
+            make_image_tensor,
+            make_image_pil,
+            make_image,
+            pytest.param(
+                make_image_cvcuda, marks=pytest.mark.skipif(not CVCUDA_AVAILABLE, reason="CVCUDA not available")
+            ),
+        ],
+    )
     def test_functional(self, make_input):
         check_functional(F.grayscale_to_rgb, make_input())
 
@@ -6483,20 +6533,50 @@ def test_functional(self, make_input):
             (F.rgb_to_grayscale_image, torch.Tensor),
             (F._color._rgb_to_grayscale_image_pil, PIL.Image.Image),
             (F.rgb_to_grayscale_image, tv_tensors.Image),
+            pytest.param(
+                F._color._rgb_to_grayscale_image_cvcuda,
+                None,
+                marks=pytest.mark.skipif(not CVCUDA_AVAILABLE, reason="CVCUDA not available"),
+            ),
         ],
     )
     def test_functional_signature(self, kernel, input_type):
+        if kernel is F._color._rgb_to_grayscale_image_cvcuda:
+            input_type = _import_cvcuda().Tensor
         check_functional_kernel_signature_match(F.grayscale_to_rgb, kernel=kernel, input_type=input_type)
 
-    @pytest.mark.parametrize("make_input", [make_image_tensor, make_image_pil, make_image])
+    @pytest.mark.parametrize(
+        "make_input",
+        [
+            make_image_tensor,
+            make_image_pil,
+            make_image,
+            pytest.param(
+                make_image_cvcuda, marks=pytest.mark.skipif(not CVCUDA_AVAILABLE, reason="CVCUDA not available")
+            ),
+        ],
+    )
     def test_transform(self, make_input):
         check_transform(transforms.RGB(), make_input(color_space="GRAY"))
 
+    @pytest.mark.parametrize(
+        "make_input",
+        [
+            make_image,
+            pytest.param(
+                make_image_cvcuda, marks=pytest.mark.skipif(not CVCUDA_AVAILABLE, reason="CVCUDA not available")
+            ),
+        ],
+    )
     @pytest.mark.parametrize("fn", [F.grayscale_to_rgb, transform_cls_to_functional(transforms.RGB)])
-    def test_image_correctness(self, fn):
-        image = make_image(dtype=torch.uint8, device="cpu", color_space="GRAY")
+    def test_image_correctness(self, make_input, fn):
+        image = make_input(dtype=torch.uint8, device="cpu", color_space="GRAY")
 
         actual = fn(image)
+
+        if make_input is make_image_cvcuda:
+            image = F.cvcuda_to_tensor(image)[0].cpu()
+
         expected = F.to_image(F.grayscale_to_rgb(F.to_pil_image(image)))
 
         assert_equal(actual, expected, rtol=0, atol=1)

torchvision/transforms/v2/_color.py

@@ -5,6 +5,7 @@
 import torch
 from torchvision import transforms as _transforms
 from torchvision.transforms.v2 import functional as F, Transform
+from torchvision.transforms.v2.functional._utils import _is_cvcuda_tensor
 
 from ._transform import _RandomApplyTransform
 from ._utils import query_chw
@@ -22,6 +23,8 @@ class Grayscale(Transform):
 
     _v1_transform_cls = _transforms.Grayscale
 
+    _transformed_types = Transform._transformed_types + (_is_cvcuda_tensor,)
+
     def __init__(self, num_output_channels: int = 1):
         super().__init__()
         self.num_output_channels = num_output_channels
@@ -44,6 +47,8 @@ class RandomGrayscale(_RandomApplyTransform):
 
     _v1_transform_cls = _transforms.RandomGrayscale
 
+    _transformed_types = _RandomApplyTransform._transformed_types + (_is_cvcuda_tensor,)
+
     def __init__(self, p: float = 0.1) -> None:
         super().__init__(p=p)
 
@@ -62,6 +67,8 @@ class RGB(Transform):
     to have [..., 1 or 3, H, W] shape, where ... means an arbitrary number of leading dimensions
     """
 
+    _transformed_types = Transform._transformed_types + (_is_cvcuda_tensor,)
+
     def __init__(self):
         super().__init__()
 

torchvision/transforms/v2/_utils.py

@@ -16,7 +16,7 @@
 
 from torchvision.transforms.transforms import _check_sequence_input, _setup_angle, _setup_size  # noqa: F401
 from torchvision.transforms.v2.functional import get_dimensions, get_size, is_pure_tensor
-from torchvision.transforms.v2.functional._utils import _FillType, _FillTypeJIT
+from torchvision.transforms.v2.functional._utils import _FillType, _FillTypeJIT, _is_cvcuda_tensor
 
 
 def _setup_number_or_seq(arg: int | float | Sequence[int | float], name: str) -> Sequence[float]:
@@ -182,7 +182,7 @@ def query_chw(flat_inputs: list[Any]) -> tuple[int, int, int]:
     chws = {
         tuple(get_dimensions(inpt))
         for inpt in flat_inputs
-        if check_type(inpt, (is_pure_tensor, tv_tensors.Image, PIL.Image.Image, tv_tensors.Video))
+        if check_type(inpt, (is_pure_tensor, tv_tensors.Image, PIL.Image.Image, tv_tensors.Video, _is_cvcuda_tensor))
     }
     if not chws:
         raise TypeError("No image or video was found in the sample")

torchvision/transforms/v2/functional/_color.py

@@ -1,3 +1,5 @@
+from typing import TYPE_CHECKING
+
 import PIL.Image
 import torch
 from torch.nn.functional import conv2d
@@ -9,7 +11,13 @@
 
 from ._misc import _num_value_bits, to_dtype_image
 from ._type_conversion import pil_to_tensor, to_pil_image
-from ._utils import _get_kernel, _register_kernel_internal
+from ._utils import _get_kernel, _import_cvcuda, _is_cvcuda_available, _register_kernel_internal
+
+
+CVCUDA_AVAILABLE = _is_cvcuda_available()
+
+if TYPE_CHECKING:
+    import cvcuda  # type: ignore[import-not-found]
 
 
 def rgb_to_grayscale(inpt: torch.Tensor, num_output_channels: int = 1) -> torch.Tensor:
@@ -63,6 +71,35 @@ def _rgb_to_grayscale_image_pil(image: PIL.Image.Image, num_output_channels: int
     return _FP.to_grayscale(image, num_output_channels=num_output_channels)
 
 
+def _rgb_to_grayscale_image_cvcuda(
+    image: "cvcuda.Tensor",
+    num_output_channels: int = 1,
+) -> "cvcuda.Tensor":
+    cvcuda = _import_cvcuda()
+
+    if not (num_output_channels == 1 or num_output_channels == 3):
+        raise ValueError(f"num_output_channels must be 1 or 3, got {num_output_channels}.")
+
+    if image.shape[3] == 1 and num_output_channels == 1:
+        # no work to do if already a single channel
+        return image
+
+    if image.shape[3] == 1 and num_output_channels == 3:
+        # just duplicate the channels
+        return cvcuda.cvtcolor(image, cvcuda.ColorConversion.GRAY2RGB)
+
+    gray = cvcuda.cvtcolor(image, cvcuda.ColorConversion.RGB2GRAY)
+
+    if num_output_channels == 3:
+        gray = cvcuda.cvtcolor(gray, cvcuda.ColorConversion.GRAY2RGB)
+
+    return gray
+
+
+if CVCUDA_AVAILABLE:
+    _register_kernel_internal(rgb_to_grayscale, _import_cvcuda().Tensor)(_rgb_to_grayscale_image_cvcuda)
+
+
 def grayscale_to_rgb(inpt: torch.Tensor) -> torch.Tensor:
     """See :class:`~torchvision.transforms.v2.RGB` for details."""
     if torch.jit.is_scripting():
@@ -89,6 +126,22 @@ def grayscale_to_rgb_image_pil(image: PIL.Image.Image) -> PIL.Image.Image:
     return image.convert(mode="RGB")
 
 
+def _grayscale_to_rgb_image_cvcuda(
+    image: "cvcuda.Tensor",
+) -> "cvcuda.Tensor":
+    cvcuda = _import_cvcuda()
+
+    if image.shape[3] == 3:
+        # if we already have RGB channels, just return the image
+        return image
+
+    return cvcuda.cvtcolor(image, cvcuda.ColorConversion.GRAY2RGB)
+
+
+if CVCUDA_AVAILABLE:
+    _register_kernel_internal(grayscale_to_rgb, _import_cvcuda().Tensor)(_grayscale_to_rgb_image_cvcuda)
+
+
 def _blend(image1: torch.Tensor, image2: torch.Tensor, ratio: float) -> torch.Tensor:
     ratio = float(ratio)
     fp = image1.is_floating_point()

torchvision/transforms/v2/functional/_meta.py

@@ -51,6 +51,14 @@ def get_dimensions_video(video: torch.Tensor) -> list[int]:
     return get_dimensions_image(video)
 
 
+def _get_dimensions_image_cvcuda(image: "cvcuda.Tensor") -> list[int]:
+    return [image.shape[3], image.shape[1], image.shape[2]]
+
+
+if CVCUDA_AVAILABLE:
+    _register_kernel_internal(get_dimensions, _import_cvcuda().Tensor)(_get_dimensions_image_cvcuda)
+
+
 def get_num_channels(inpt: torch.Tensor) -> int:
     if torch.jit.is_scripting():
         return get_num_channels_image(inpt)