Equalize 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
@@ -5449,7 +5449,18 @@ def test_kernel_image(self, dtype, device):
     def test_kernel_video(self):
         check_kernel(F.equalize_image, make_video())
 
-    @pytest.mark.parametrize("make_input", [make_image_tensor, make_image_pil, make_image, make_video])
+    @pytest.mark.parametrize(
+        "make_input",
+        [
+            make_image_tensor,
+            make_image_pil,
+            make_image,
+            make_video,
+            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.equalize, make_input())
 
@@ -5460,33 +5471,68 @@ def test_functional(self, make_input):
             (F._color._equalize_image_pil, PIL.Image.Image),
             (F.equalize_image, tv_tensors.Image),
             (F.equalize_video, tv_tensors.Video),
+            pytest.param(
+                F._color._equalize_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._equalize_image_cvcuda:
+            input_type = _import_cvcuda().Tensor
         check_functional_kernel_signature_match(F.equalize, kernel=kernel, input_type=input_type)
 
     @pytest.mark.parametrize(
         "make_input",
-        [make_image_tensor, make_image_pil, make_image, make_video],
+        [
+            make_image_tensor,
+            make_image_pil,
+            make_image,
+            make_video,
+            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.RandomEqualize(p=1), make_input())
 
     @pytest.mark.parametrize(("low", "high"), [(0, 64), (64, 192), (192, 256), (0, 1), (127, 128), (255, 256)])
+    @pytest.mark.parametrize(
+        "tensor_type",
+        [
+            torch.Tensor,
+            pytest.param(
+                "cvcuda.Tensor", marks=pytest.mark.skipif(not CVCUDA_AVAILABLE, reason="CVCUDA not available")
+            ),
+        ],
+    )
     @pytest.mark.parametrize("fn", [F.equalize, transform_cls_to_functional(transforms.RandomEqualize, p=1)])
-    def test_image_correctness(self, low, high, fn):
+    def test_image_correctness(self, low, high, tensor_type, fn):
         # We are not using the default `make_image` here since that uniformly samples the values over the whole value
         # range. Since the whole point of F.equalize is to transform an arbitrary distribution of values into a uniform
         # one over the full range, the information gain is low if we already provide something really close to the
         # expected value.
-        image = tv_tensors.Image(
-            torch.testing.make_tensor((3, 117, 253), dtype=torch.uint8, device="cpu", low=low, high=high)
-        )
+        shape = (3, 117, 253)
+        if tensor_type == "cvcuda.Tensor":
+            shape = (1, *shape)
+        image = tv_tensors.Image(torch.testing.make_tensor(shape, dtype=torch.uint8, device="cpu", low=low, high=high))
+
+        if tensor_type == "cvcuda.Tensor":
+            image = F.to_cvcuda_tensor(image)
 
         actual = fn(image)
+
+        if tensor_type == "cvcuda.Tensor":
+            image = F.cvcuda_to_tensor(image)[0].cpu()
+
         expected = F.to_image(F.equalize(F.to_pil_image(image)))
 
-        assert_equal(actual, expected)
+        if tensor_type == "cvcuda.Tensor":
+            assert_close(actual, expected, rtol=1e-10, atol=1)
+        else:
+            assert_equal(actual, expected)
 
 
 class TestUniformTemporalSubsample:

torchvision/transforms/v2/_color.py

@@ -5,11 +5,15 @@
 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_available, _is_cvcuda_tensor
 
 from ._transform import _RandomApplyTransform
 from ._utils import query_chw
 
 
+CVCUDA_AVAILABLE = _is_cvcuda_available()
+
+
 class Grayscale(Transform):
     """Convert images or videos to grayscale.
 
@@ -265,6 +269,9 @@ class RandomEqualize(_RandomApplyTransform):
 
     _v1_transform_cls = _transforms.RandomEqualize
 
+    if CVCUDA_AVAILABLE:
+        _transformed_types = _RandomApplyTransform._transformed_types + (_is_cvcuda_tensor,)
+
     def transform(self, inpt: Any, params: dict[str, Any]) -> Any:
         return self._call_kernel(F.equalize, inpt)
 

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")
@@ -207,6 +207,7 @@ def query_size(flat_inputs: list[Any]) -> tuple[int, int]:
                 tv_tensors.Mask,
                 tv_tensors.BoundingBoxes,
                 tv_tensors.KeyPoints,
+                _is_cvcuda_tensor,
             ),
         )
     }

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 +11,15 @@
 
 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]
+if CVCUDA_AVAILABLE:
+    cvcuda = _import_cvcuda()  # noqa: F811
 
 
 def rgb_to_grayscale(inpt: torch.Tensor, num_output_channels: int = 1) -> torch.Tensor:
@@ -649,6 +659,17 @@ def equalize_video(video: torch.Tensor) -> torch.Tensor:
     return equalize_image(video)
 
 
+def _equalize_image_cvcuda(
+    image: "cvcuda.Tensor",
+) -> "cvcuda.Tensor":
+    cvcuda = _import_cvcuda()
+    return cvcuda.histogrameq(image, dtype=image.dtype)
+
+
+if CVCUDA_AVAILABLE:
+    _register_kernel_internal(equalize, _import_cvcuda().Tensor)(_equalize_image_cvcuda)
+
+
 def invert(inpt: torch.Tensor) -> torch.Tensor:
     """See :func:`~torchvision.transforms.v2.RandomInvert`."""
     if torch.jit.is_scripting():

torchvision/transforms/v2/functional/_meta.py

@@ -51,6 +51,16 @@ def get_dimensions_video(video: torch.Tensor) -> list[int]:
     return get_dimensions_image(video)
 
 
+def get_dimensions_image_cvcuda(image: "cvcuda.Tensor") -> list[int]:
+    # CV-CUDA tensor is always in NHWC layout
+    # get_dimensions is CHW
+    return [image.shape[3], image.shape[1], image.shape[2]]
+
+
+if CVCUDA_AVAILABLE:
+    _register_kernel_internal(get_dimensions, 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)
@@ -87,6 +97,16 @@ def get_num_channels_video(video: torch.Tensor) -> int:
 get_image_num_channels = get_num_channels
 
 
+def get_num_channels_image_cvcuda(image: "cvcuda.Tensor") -> int:
+    # CV-CUDA tensor is always in NHWC layout
+    # get_num_channels is C
+    return image.shape[3]
+
+
+if CVCUDA_AVAILABLE:
+    _register_kernel_internal(get_num_channels, cvcuda.Tensor)(get_num_channels_image_cvcuda)
+
+
 def get_size(inpt: torch.Tensor) -> list[int]:
     if torch.jit.is_scripting():
         return get_size_image(inpt)
@@ -125,7 +145,7 @@ def get_size_image_cvcuda(image: "cvcuda.Tensor") -> list[int]:
 
 
 if CVCUDA_AVAILABLE:
-    _get_size_image_cvcuda = _register_kernel_internal(get_size, cvcuda.Tensor)(get_size_image_cvcuda)
+    _register_kernel_internal(get_size, _import_cvcuda().Tensor)(get_size_image_cvcuda)
 
 
 @_register_kernel_internal(get_size, tv_tensors.Video, tv_tensor_wrapper=False)