test/morphology/test_gradient.py

import pytest
import torch
from torch.autograd import gradcheck

from kornia.morphology import gradient
from kornia.testing import assert_close


class TestGradient:
    def test_smoke(self, device, dtype):
        kernel = torch.rand(3, 3, device=device, dtype=dtype)
        assert kernel is not None

    @pytest.mark.parametrize("shape", [(1, 3, 4, 4), (2, 3, 2, 4), (3, 3, 4, 1), (3, 2, 5, 5)])
    @pytest.mark.parametrize("kernel", [(3, 3), (5, 5)])
    def test_cardinality(self, device, dtype, shape, kernel):
        img = torch.ones(shape, device=device, dtype=dtype)
        krnl = torch.ones(kernel, device=device, dtype=dtype)
        assert gradient(img, krnl).shape == shape

    def test_kernel(self, device, dtype):
        tensor = torch.tensor([[0.5, 1.0, 0.3], [0.7, 0.3, 0.8], [0.4, 0.9, 0.2]], device=device, dtype=dtype)[
            None, None, :, :
        ]
        kernel = torch.tensor([[0.0, 1.0, 0.0], [1.0, 1.0, 1.0], [0.0, 1.0, 0.0]], device=device, dtype=dtype)
        expected = torch.tensor([[0.5, 0.7, 0.7], [0.4, 0.7, 0.6], [0.5, 0.7, 0.7]], device=device, dtype=dtype)[
            None, None, :, :
        ]
        assert_close(gradient(tensor, kernel), expected, atol=1e-3, rtol=1e-3)

    def test_structural_element(self, device, dtype):
        tensor = torch.tensor([[0.5, 1.0, 0.3], [0.7, 0.3, 0.8], [0.4, 0.9, 0.2]], device=device, dtype=dtype)[
            None, None, :, :
        ]
        structural_element = torch.tensor(
            [[-1.0, 0.0, -1.0], [0.0, 0.0, 0.0], [-1.0, 0.0, -1.0]], device=device, dtype=dtype
        )
        expected = torch.tensor([[0.5, 0.7, 0.7], [0.4, 0.7, 0.6], [0.5, 0.7, 0.7]], device=device, dtype=dtype)[
            None, None, :, :
        ]
        assert_close(
            gradient(tensor, torch.ones_like(structural_element), structuring_element=structural_element),
            expected,
            atol=1e-3,
            rtol=1e-3,
        )

    def test_exception(self, device, dtype):
        tensor = torch.ones(1, 1, 3, 4, device=device, dtype=dtype)
        kernel = torch.ones(3, 3, device=device, dtype=dtype)

        with pytest.raises(TypeError):
            assert gradient([0.0], kernel)

        with pytest.raises(TypeError):
            assert gradient(tensor, [0.0])

        with pytest.raises(ValueError):
            test = torch.ones(2, 3, 4, device=device, dtype=dtype)
            assert gradient(test, kernel)

        with pytest.raises(ValueError):
            test = torch.ones(2, 3, 4, device=device, dtype=dtype)
            assert gradient(tensor, test)

    @pytest.mark.grad
    def test_gradcheck(self, device, dtype):
        tensor = torch.rand(2, 3, 4, 4, requires_grad=True, device=device, dtype=torch.float64)
        kernel = torch.rand(3, 3, requires_grad=True, device=device, dtype=torch.float64)
        assert gradcheck(gradient, (tensor, kernel), raise_exception=True)

    @pytest.mark.jit
    def test_jit(self, device, dtype):
        op = gradient
        op_script = torch.jit.script(op)

        tensor = torch.rand(1, 2, 7, 7, device=device, dtype=dtype)
        kernel = torch.ones(3, 3, device=device, dtype=dtype)

        actual = op_script(tensor, kernel)
        expected = op(tensor, kernel)

        assert_close(actual, expected)