File size: 10,055 Bytes
36c95ba |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 |
import pytest
import torch
from torch.autograd import gradcheck
import kornia
from kornia.testing import assert_close, BaseTester
class TestRgbToYcbcr(BaseTester):
def test_smoke(self, device, dtype):
C, H, W = 3, 4, 5
img = torch.rand(C, H, W, device=device, dtype=dtype)
assert isinstance(kornia.color.rgb_to_ycbcr(img), torch.Tensor)
@pytest.mark.parametrize("shape", [(1, 3, 4, 4), (2, 3, 2, 4), (3, 3, 4, 1), (3, 2, 1)])
def test_cardinality(self, device, dtype, shape):
img = torch.ones(shape, device=device, dtype=dtype)
assert kornia.color.rgb_to_ycbcr(img).shape == shape
def test_exception(self, device, dtype):
with pytest.raises(TypeError):
assert kornia.color.rgb_to_ycbcr([0.0])
with pytest.raises(ValueError):
img = torch.ones(1, 1, device=device, dtype=dtype)
assert kornia.color.rgb_to_ycbcr(img)
with pytest.raises(ValueError):
img = torch.ones(2, 1, 1, device=device, dtype=dtype)
assert kornia.color.rgb_to_ycbcr(img)
def test_unit(self, device, dtype):
data = torch.tensor(
[
[
[
[0.7925, 0.8704, 0.2771, 0.3279, 0.0193],
[0.2483, 0.6212, 0.9859, 0.5044, 0.5621],
[0.5762, 0.3959, 0.2931, 0.2669, 0.0243],
[0.6989, 0.0529, 0.8344, 0.6523, 0.8980],
[0.5181, 0.9341, 0.2172, 0.0520, 0.7266],
],
[
[0.8413, 0.0284, 0.3625, 0.8864, 0.5595],
[0.3791, 0.0235, 0.4251, 0.0619, 0.5270],
[0.3516, 0.8005, 0.9571, 0.4113, 0.6119],
[0.0632, 0.8836, 0.0261, 0.1550, 0.4923],
[0.2332, 0.7044, 0.9514, 0.2443, 0.2818],
],
[
[0.6899, 0.2063, 0.3179, 0.8989, 0.4378],
[0.0384, 0.5230, 0.6416, 0.9749, 0.7863],
[0.8577, 0.3115, 0.2375, 0.5446, 0.9837],
[0.3213, 0.6618, 0.5977, 0.3999, 0.4962],
[0.1385, 0.5831, 0.9756, 0.8714, 0.8017],
],
]
],
device=device,
dtype=dtype,
)
# Reference output generated using OpenCV: cv2.cvtColor(data, cv2.COLOR_RGB2XYZ)
expected = torch.tensor(
[
[
[
[0.8094629, 0.30041942, 0.33188093, 0.7208117, 0.38409105],
[0.30114722, 0.2591345, 0.6174494, 0.29830942, 0.5670594],
[0.47643244, 0.623786, 0.6765313, 0.38332558, 0.47857922],
[0.2826805, 0.60994685, 0.33298355, 0.3316514, 0.6140681],
[0.30756146, 0.75926465, 0.7346588, 0.25828302, 0.4740529],
],
[
[0.43258172, 0.4469004, 0.492101, 0.60042214, 0.53028214],
[0.35180065, 0.64882994, 0.51362985, 0.8815981, 0.62366176],
[0.71505237, 0.32387614, 0.2523859, 0.5909421, 0.78490996],
[0.52178127, 0.529264, 0.64931786, 0.53848785, 0.43353173],
[0.4046721, 0.40065458, 0.6358658, 0.8458253, 0.6847739],
],
[
[0.48791525, 0.9063825, 0.46094114, 0.21983841, 0.23989305],
[0.46231723, 0.7581379, 0.7626976, 0.6469568, 0.496467],
[0.5711212, 0.33752257, 0.22661471, 0.4169921, 0.17608929],
[0.7967522, 0.10283372, 0.85753804, 0.72865105, 0.70245713],
[0.65009415, 0.6246666, 0.13107029, 0.35291404, 0.6800583],
],
]
],
device=device,
dtype=dtype,
)
assert_close(kornia.color.rgb_to_ycbcr(data), expected, atol=1e-4, rtol=1e-4)
# TODO: investigate and implement me
# def test_forth_and_back(self, device, dtype):
# pass
@pytest.mark.grad
def test_gradcheck(self, device, dtype):
B, C, H, W = 2, 3, 4, 4
img = torch.rand(B, C, H, W, device=device, dtype=torch.float64, requires_grad=True)
assert gradcheck(kornia.color.rgb_to_ycbcr, (img,), raise_exception=True)
@pytest.mark.jit
def test_jit(self, device, dtype):
B, C, H, W = 2, 3, 4, 4
img = torch.ones(B, C, H, W, device=device, dtype=dtype)
op = kornia.color.rgb_to_ycbcr
op_jit = torch.jit.script(op)
assert_close(op(img), op_jit(img))
@pytest.mark.nn
def test_module(self, device, dtype):
B, C, H, W = 2, 3, 4, 4
img = torch.ones(B, C, H, W, device=device, dtype=dtype)
ops = kornia.color.RgbToYcbcr().to(device, dtype)
fcn = kornia.color.rgb_to_ycbcr
assert_close(ops(img), fcn(img))
class TestYcbcrToRgb(BaseTester):
def test_smoke(self, device, dtype):
C, H, W = 3, 4, 5
img = torch.rand(C, H, W, device=device, dtype=dtype)
assert isinstance(kornia.color.ycbcr_to_rgb(img), torch.Tensor)
@pytest.mark.parametrize("shape", [(1, 3, 4, 4), (2, 3, 2, 4), (3, 3, 4, 1), (3, 2, 1)])
def test_cardinality(self, device, dtype, shape):
img = torch.ones(shape, device=device, dtype=dtype)
assert kornia.color.ycbcr_to_rgb(img).shape == shape
def test_exception(self, device, dtype):
with pytest.raises(TypeError):
assert kornia.color.ycbcr_to_rgb([0.0])
with pytest.raises(ValueError):
img = torch.ones(1, 1, device=device, dtype=dtype)
assert kornia.color.ycbcr_to_rgb(img)
with pytest.raises(ValueError):
img = torch.ones(2, 1, 1, device=device, dtype=dtype)
assert kornia.color.ycbcr_to_rgb(img)
def test_unit(self, device, dtype):
data = torch.tensor(
[
[
[
[0.9892, 0.2620, 0.4184, 0.5286, 0.2793],
[0.0722, 0.8828, 0.8714, 0.0657, 0.7798],
[0.8118, 0.7522, 0.0260, 0.8811, 0.5226],
[0.0644, 0.3648, 0.4448, 0.4202, 0.7316],
[0.9138, 0.1956, 0.4257, 0.6381, 0.1353],
],
[
[0.7408, 0.8529, 0.5119, 0.0220, 0.0226],
[0.8963, 0.5652, 0.9568, 0.6977, 0.8221],
[0.4645, 0.0478, 0.4952, 0.5492, 0.4861],
[0.9980, 0.9978, 0.0281, 0.5283, 0.8146],
[0.7789, 0.2663, 0.6437, 0.6926, 0.5627],
],
[
[0.7377, 0.7152, 0.3080, 0.8515, 0.4841],
[0.7192, 0.3297, 0.7337, 0.0230, 0.2464],
[0.6399, 0.8998, 0.3838, 0.3043, 0.3774],
[0.1281, 0.6731, 0.4218, 0.3963, 0.8541],
[0.2245, 0.2413, 0.2351, 0.9522, 0.8158],
],
]
],
device=device,
dtype=dtype,
)
# Reference output generated using OpenCV: cv2.cvtColor(data, cv2.COLOR_RGB2XYZ)
expected = torch.tensor(
[
[
[
[1.3226931, 0.5639256, 0.14902398, 1.0217545, 0.2569923],
[0.37973762, 0.64386904, 1.1992811, -0.603531, 0.4239992],
[1.0080798, 1.3131194, -0.1370286, 0.60653293, 0.3505922],
[-0.4573757, 0.6076593, 0.33508536, 0.27470887, 1.2284023],
[0.52727354, -0.1673561, 0.05404532, 1.2725366, 0.5783674],
],
[
[0.736647, -0.01305042, 0.55139434, 0.44206098, 0.4548782],
[-0.22063601, 0.98196536, 0.54739904, 0.33826917, 0.850068],
[0.72412336, 0.6222996, 0.110618, 1.0039049, 0.614918],
[0.15862459, 0.0699634, 0.66296846, 0.4845066, 0.3705502],
[1.0145653, 0.46070462, 0.5654058, 0.24897486, -0.11174999],
],
[
[1.4161384, 0.88769174, 0.4394987, -0.31889397, -0.5671302],
[0.77483994, 0.99839956, 1.6813064, 0.41622213, 1.3508832],
[0.7488585, -0.04955059, 0.01748962, 0.9683316, 0.49795526],
[0.9473541, 1.2473994, -0.3918787, 0.47037587, 1.2893858],
[1.4082898, -0.21875012, 0.6804801, 0.9795798, 0.24646705],
],
]
],
device=device,
dtype=dtype,
)
assert_close(kornia.color.ycbcr_to_rgb(data), expected, atol=1e-4, rtol=1e-4)
# TODO: investigate and implement me
# def test_forth_and_back(self, device, dtype):
# pass
@pytest.mark.grad
def test_gradcheck(self, device, dtype):
B, C, H, W = 2, 3, 4, 4
img = torch.rand(B, C, H, W, device=device, dtype=torch.float64, requires_grad=True)
assert gradcheck(kornia.color.ycbcr_to_rgb, (img,), raise_exception=True)
@pytest.mark.jit
def test_jit(self, device, dtype):
B, C, H, W = 2, 3, 4, 4
img = torch.ones(B, C, H, W, device=device, dtype=dtype)
op = kornia.color.ycbcr_to_rgb
op_jit = torch.jit.script(op)
assert_close(op(img), op_jit(img))
@pytest.mark.nn
def test_module(self, device, dtype):
B, C, H, W = 2, 3, 4, 4
img = torch.ones(B, C, H, W, device=device, dtype=dtype)
ops = kornia.color.YcbcrToRgb().to(device, dtype)
fcn = kornia.color.ycbcr_to_rgb
assert_close(ops(img), fcn(img))
|