From 97aae46b66d4b4612df00a13049714cc2a293548 Mon Sep 17 00:00:00 2001
From: xuan07472 <xuan07472@gmail.com>
Date: Thu, 12 Jan 2023 21:04:36 +0800
Subject: [PATCH 01/16] add gradient error

---
 docs/en/api/metrics.rst                   |   1 +
 docs/zh_cn/api/metrics.rst                |   1 +
 mmeval/metrics/__init__.py                |   3 +-
 mmeval/metrics/gradient_error.py          | 176 ++++++++++++++++++++++
 tests/test_metrics/test_gradient_error.py |  19 +++
 5 files changed, 199 insertions(+), 1 deletion(-)
 create mode 100644 mmeval/metrics/gradient_error.py
 create mode 100644 tests/test_metrics/test_gradient_error.py

diff --git a/docs/en/api/metrics.rst b/docs/en/api/metrics.rst
index a7a275d7..16ee6f01 100644
--- a/docs/en/api/metrics.rst
+++ b/docs/en/api/metrics.rst
@@ -42,3 +42,4 @@ Metrics
    MAE
    MSE
    BLEU
+   GradientError
diff --git a/docs/zh_cn/api/metrics.rst b/docs/zh_cn/api/metrics.rst
index a7a275d7..16ee6f01 100644
--- a/docs/zh_cn/api/metrics.rst
+++ b/docs/zh_cn/api/metrics.rst
@@ -42,3 +42,4 @@ Metrics
    MAE
    MSE
    BLEU
+   GradientError
diff --git a/mmeval/metrics/__init__.py b/mmeval/metrics/__init__.py
index 31d2b5bd..1b973299 100644
--- a/mmeval/metrics/__init__.py
+++ b/mmeval/metrics/__init__.py
@@ -6,6 +6,7 @@
 from .coco_detection import COCODetectionMetric
 from .end_point_error import EndPointError
 from .f_metric import F1Metric
+from .gradient_error import GradientError
 from .hmean_iou import HmeanIoU
 from .mae import MAE
 from .mean_iou import MeanIoU
@@ -25,5 +26,5 @@
     'F1Metric', 'HmeanIoU', 'SingleLabelMetric', 'COCODetectionMetric',
     'PCKAccuracy', 'MpiiPCKAccuracy', 'JhmdbPCKAccuracy', 'ProposalRecall',
     'PSNR', 'MAE', 'MSE', 'SSIM', 'SNR', 'MultiLabelMetric',
-    'AveragePrecision', 'AVAMeanAP', 'BLEU'
+    'AveragePrecision', 'AVAMeanAP', 'BLEU', 'GradientError'
 ]
diff --git a/mmeval/metrics/gradient_error.py b/mmeval/metrics/gradient_error.py
new file mode 100644
index 00000000..6e8ec9ba
--- /dev/null
+++ b/mmeval/metrics/gradient_error.py
@@ -0,0 +1,176 @@
+# Copyright (c) OpenMMLab. All rights reserved.
+import cv2
+import numpy as np
+from typing import Dict, List, Sequence
+
+from mmeval.core import BaseMetric
+
+
+def gaussian(x, sigma):
+    """Gaussian function.
+
+    Args:
+        x (array_like): The independent variable.
+        sigma (float): Standard deviation of the gaussian function.
+
+    Return:
+        np.ndarray or scalar: Gaussian value of `x`.
+    """
+    return np.exp(-x**2 / (2 * sigma**2)) / (sigma * np.sqrt(2 * np.pi))
+
+
+def dgaussian(x, sigma):
+    """Gradient of gaussian.
+
+    Args:
+        x (array_like): The independent variable.
+        sigma (float): Standard deviation of the gaussian function.
+
+    Return:
+        np.ndarray or scalar: Gradient of gaussian of `x`.
+    """
+    return -x * gaussian(x, sigma) / sigma**2
+
+
+def gauss_filter(sigma, epsilon=1e-2):
+    """Gradient of gaussian.
+
+    Args:
+        sigma (float): Standard deviation of the gaussian kernel.
+        epsilon (float): Small value used when calculating kernel size.
+            Default: 1e-2.
+
+    Return:
+        filter_x (np.ndarray): Gaussian filter along x axis.
+        filter_y (np.ndarray): Gaussian filter along y axis.
+    """
+    half_size = np.ceil(
+        sigma * np.sqrt(-2 * np.log(np.sqrt(2 * np.pi) * sigma * epsilon)))
+    size = int(2 * half_size + 1)
+
+    # create filter in x axis
+    filter_x = np.zeros((size, size))
+    for i in range(size):
+        for j in range(size):
+            filter_x[i, j] = gaussian(i - half_size, sigma) * dgaussian(
+                j - half_size, sigma)
+
+    # normalize filter
+    norm = np.sqrt((filter_x**2).sum())
+    filter_x = filter_x / norm
+    filter_y = np.transpose(filter_x)
+
+    return filter_x, filter_y
+
+
+def gauss_gradient(img, sigma):
+    """Gaussian gradient.
+
+    From https://www.mathworks.com/matlabcentral/mlc-downloads/downloads/
+    submissions/8060/versions/2/previews/gaussgradient/gaussgradient.m/
+    index.html
+
+    Args:
+        img (np.ndarray): Input image.
+        sigma (float): Standard deviation of the gaussian kernel.
+
+    Return:
+        np.ndarray: Gaussian gradient of input `img`.
+    """
+    filter_x, filter_y = gauss_filter(sigma)
+    img_filtered_x = cv2.filter2D(
+        img, -1, filter_x, borderType=cv2.BORDER_REPLICATE)
+    img_filtered_y = cv2.filter2D(
+        img, -1, filter_y, borderType=cv2.BORDER_REPLICATE)
+    return np.sqrt(img_filtered_x**2 + img_filtered_y**2)
+
+
+class GradientError(BaseMetric):
+    """Gradient error for evaluating alpha matte prediction.
+
+    Args:
+        sigma (float): Standard deviation of the gaussian kernel.
+            Defaults to 1.4 .
+        norm_const (int): Divide the result to reduce its magnitude.
+            Defaults to 1000 .
+        **kwargs: Keyword parameters passed to :class:`BaseMetric`.
+    Note:
+        The current implementation assumes the image / alpha / trimap
+        a numpy array with pixel values ranging from 0 to 255.
+
+        The pred_alpha should be masked by trimap before passing
+        into this metric.
+
+        The trimap is the most commonly used prior knowledge. As the
+        name implies, trimap is a ternary graph and each pixel
+        takes one of {0, 128, 255}, representing the foreground, the
+        unknown and the background respectively.
+
+    Examples:
+        >>> from mmeval import GradientError
+        >>> import numpy as np
+        >>>
+        >>> gradienterror = GradientError()
+        >>> pred_alpha = np.zeros((32, 32), dtype=np.uint8)
+        >>> gt_alpha = np.ones((32, 32), dtype=np.uint8) * 255
+        >>> trimap = np.zeros((32, 32), dtype=np.uint8)
+        >>> trimap[:16, :16] = 128
+        >>> trimap[16:, 16:] = 255
+        >>> gradienterror(pred_alpha, gt_alpha, trimap)  # doctest: +ELLIPSIS
+        {'GradientError': ...}
+    """
+
+    def __init__(self,
+                 sigma: float = 1.4,
+                 norm_const: int = 1000,
+                 **kwargs) -> None:
+        super().__init__(**kwargs)
+        self.sigma = sigma
+        self.norm_const = norm_const
+
+    def add(self, pred_alphas: Sequence[np.ndarray], gt_alphas: Sequence[np.ndarray], trimaps: Sequence[np.ndarray]) -> None:  # type: ignore # yapf: disable # noqa: E501
+        """Add GradientError score of batch to ``self._results``
+
+        Args:
+            pred_alpha(Sequence[np.ndarray]): Pred_alpha data of predictions.
+            ori_alpha(Sequence[np.ndarray]): Ori_alpha data of data_batch.
+            ori_trimap(Sequence[np.ndarray]): Ori_trimap data of data_batch.
+        """
+
+        for pred_alpha, gt_alpha, trimap in zip(pred_alphas, gt_alphas,
+                                                trimaps):
+            assert pred_alpha.shape == gt_alpha.shape, 'The shape of ' \
+                '`pred_alpha` and `gt_alpha` should be the same, but got: ' \
+                f'{pred_alpha.shape} and {gt_alpha.shape}'
+
+            gt_alpha_normed = np.zeros_like(gt_alpha)
+            pred_alpha_normed = np.zeros_like(pred_alpha)
+
+            cv2.normalize(gt_alpha, gt_alpha_normed, 1.0, 0.0, cv2.NORM_MINMAX)
+            cv2.normalize(pred_alpha, pred_alpha_normed, 1.0, 0.0,
+                          cv2.NORM_MINMAX)
+
+            gt_alpha_grad = gauss_gradient(gt_alpha_normed, self.sigma)
+            pred_alpha_grad = gauss_gradient(pred_alpha_normed, self.sigma)
+            # this is the sum over n samples
+            grad_loss = ((gt_alpha_grad - pred_alpha_grad)**2 *
+                         (trimap == 128)).sum()
+
+            # divide by 1000 to reduce the magnitude of the result
+            grad_loss /= self.norm_const
+
+            self._results.append(grad_loss)
+
+    def compute_metric(self, results: List) -> Dict[str, float]:
+        """Compute the GradientError metric.
+
+        Args:
+            results (List): A list that consisting the GradientError score.
+                This list has already been synced across all ranks.
+        Returns:
+            Dict[str, float]: The computed GradientError metric.
+            The keys are the names of the metrics,
+            and the values are corresponding results.
+        """
+
+        return {'GradientError': float(np.array(results).mean())}
diff --git a/tests/test_metrics/test_gradient_error.py b/tests/test_metrics/test_gradient_error.py
new file mode 100644
index 00000000..ab833c97
--- /dev/null
+++ b/tests/test_metrics/test_gradient_error.py
@@ -0,0 +1,19 @@
+# Copyright (c) OpenMMLab. All rights reserved.
+import numpy as np
+
+from mmeval.metrics import GradientError
+
+
+def test_gradient_error():
+    np.random.seed(0)
+    pred_alpha = np.random.randn(32, 32).astype('uint8')
+    gt_alpha = np.ones((32, 32), dtype=np.uint8) * 255
+    trimap = np.zeros((32, 32), dtype=np.uint8)
+    trimap[:16, :16] = 128
+    trimap[16:, 16:] = 255
+
+    gradienterror = GradientError()
+    gradienterror_results = gradienterror(pred_alpha, gt_alpha, trimap)
+    assert isinstance(gradienterror_results, dict)
+    np.testing.assert_almost_equal(gradienterror_results['GradientError'],
+                                   0.0935)

From 4aab935e948428a902ffcf1b1b030135bd48d039 Mon Sep 17 00:00:00 2001
From: xuan07472 <xuan07472@gmail.com>
Date: Fri, 13 Jan 2023 13:53:24 +0800
Subject: [PATCH 02/16] update docstring and type hints

---
 mmeval/metrics/gradient_error.py | 47 ++++++++++++++++++--------------
 1 file changed, 27 insertions(+), 20 deletions(-)

diff --git a/mmeval/metrics/gradient_error.py b/mmeval/metrics/gradient_error.py
index 6e8ec9ba..9c2865d6 100644
--- a/mmeval/metrics/gradient_error.py
+++ b/mmeval/metrics/gradient_error.py
@@ -6,44 +6,46 @@
 from mmeval.core import BaseMetric
 
 
-def gaussian(x, sigma):
+def gaussian(x: np.ndarray, sigma: float):
     """Gaussian function.
 
     Args:
         x (array_like): The independent variable.
         sigma (float): Standard deviation of the gaussian function.
 
-    Return:
+    Returns:
         np.ndarray or scalar: Gaussian value of `x`.
     """
+
     return np.exp(-x**2 / (2 * sigma**2)) / (sigma * np.sqrt(2 * np.pi))
 
 
-def dgaussian(x, sigma):
-    """Gradient of gaussian.
+def dgaussian(x: np.ndarray, sigma: float):
+    """Derivative of Gaussian function.
 
     Args:
         x (array_like): The independent variable.
         sigma (float): Standard deviation of the gaussian function.
 
-    Return:
+    Returns:
         np.ndarray or scalar: Gradient of gaussian of `x`.
     """
+
     return -x * gaussian(x, sigma) / sigma**2
 
 
-def gauss_filter(sigma, epsilon=1e-2):
-    """Gradient of gaussian.
+def gauss_filter(sigma: float, epsilon=1e-2):
+    """Gaussian Filter.
 
     Args:
         sigma (float): Standard deviation of the gaussian kernel.
         epsilon (float): Small value used when calculating kernel size.
             Default: 1e-2.
 
-    Return:
-        filter_x (np.ndarray): Gaussian filter along x axis.
-        filter_y (np.ndarray): Gaussian filter along y axis.
+    Returns:
+        tuple(np.ndarray, np.ndarray): Gaussian filter along x and y axis.
     """
+
     half_size = np.ceil(
         sigma * np.sqrt(-2 * np.log(np.sqrt(2 * np.pi) * sigma * epsilon)))
     size = int(2 * half_size + 1)
@@ -63,10 +65,13 @@ def gauss_filter(sigma, epsilon=1e-2):
     return filter_x, filter_y
 
 
-def gauss_gradient(img, sigma):
-    """Gaussian gradient.
+def gauss_gradient(img: np.ndarray, sigma: float):
+    """Gaussian gradient.The first order Gaussian derivative convolution
+    calculation is carried out by using Gaussian filter. Calculate their
+    gradients separately, make a difference, and then accumulate their squares.
+    The more similar the two, the smaller the Gradient error.
 
-    From https://www.mathworks.com/matlabcentral/mlc-downloads/downloads/
+    Reference: https://www.mathworks.com/matlabcentral/mlc-downloads/downloads/
     submissions/8060/versions/2/previews/gaussgradient/gaussgradient.m/
     index.html
 
@@ -74,9 +79,10 @@ def gauss_gradient(img, sigma):
         img (np.ndarray): Input image.
         sigma (float): Standard deviation of the gaussian kernel.
 
-    Return:
+    Returns:
         np.ndarray: Gaussian gradient of input `img`.
     """
+
     filter_x, filter_y = gauss_filter(sigma)
     img_filtered_x = cv2.filter2D(
         img, -1, filter_x, borderType=cv2.BORDER_REPLICATE)
@@ -110,13 +116,13 @@ class GradientError(BaseMetric):
         >>> from mmeval import GradientError
         >>> import numpy as np
         >>>
-        >>> gradienterror = GradientError()
+        >>> gradient_error = GradientError()
         >>> pred_alpha = np.zeros((32, 32), dtype=np.uint8)
         >>> gt_alpha = np.ones((32, 32), dtype=np.uint8) * 255
         >>> trimap = np.zeros((32, 32), dtype=np.uint8)
         >>> trimap[:16, :16] = 128
         >>> trimap[16:, 16:] = 255
-        >>> gradienterror(pred_alpha, gt_alpha, trimap)  # doctest: +ELLIPSIS
+        >>> gradient_error(pred_alpha, gt_alpha, trimap)  # doctest: +ELLIPSIS
         {'GradientError': ...}
     """
 
@@ -132,9 +138,9 @@ def add(self, pred_alphas: Sequence[np.ndarray], gt_alphas: Sequence[np.ndarray]
         """Add GradientError score of batch to ``self._results``
 
         Args:
-            pred_alpha(Sequence[np.ndarray]): Pred_alpha data of predictions.
-            ori_alpha(Sequence[np.ndarray]): Ori_alpha data of data_batch.
-            ori_trimap(Sequence[np.ndarray]): Ori_trimap data of data_batch.
+            pred_alphas (Sequence[np.ndarray]): Pred_alpha data of predictions.
+            ori_alphas (Sequence[np.ndarray]): Ori_alpha data of data_batch.
+            ori_trimaps (Sequence[np.ndarray]): Ori_trimap data of data_batch.
         """
 
         for pred_alpha, gt_alpha, trimap in zip(pred_alphas, gt_alphas,
@@ -156,7 +162,7 @@ def add(self, pred_alphas: Sequence[np.ndarray], gt_alphas: Sequence[np.ndarray]
             grad_loss = ((gt_alpha_grad - pred_alpha_grad)**2 *
                          (trimap == 128)).sum()
 
-            # divide by 1000 to reduce the magnitude of the result
+            # divide by self.norm_const to reduce the magnitude of the result
             grad_loss /= self.norm_const
 
             self._results.append(grad_loss)
@@ -167,6 +173,7 @@ def compute_metric(self, results: List) -> Dict[str, float]:
         Args:
             results (List): A list that consisting the GradientError score.
                 This list has already been synced across all ranks.
+
         Returns:
             Dict[str, float]: The computed GradientError metric.
             The keys are the names of the metrics,

From f2de3aeca703615139cf416490c9e35c3d87b8c1 Mon Sep 17 00:00:00 2001
From: xuan07472 <xuan07472@gmail.com>
Date: Fri, 13 Jan 2023 13:57:22 +0800
Subject: [PATCH 03/16] update

---
 mmeval/metrics/gradient_error.py | 1 +
 1 file changed, 1 insertion(+)

diff --git a/mmeval/metrics/gradient_error.py b/mmeval/metrics/gradient_error.py
index 9c2865d6..6cbe8659 100644
--- a/mmeval/metrics/gradient_error.py
+++ b/mmeval/metrics/gradient_error.py
@@ -100,6 +100,7 @@ class GradientError(BaseMetric):
         norm_const (int): Divide the result to reduce its magnitude.
             Defaults to 1000 .
         **kwargs: Keyword parameters passed to :class:`BaseMetric`.
+
     Note:
         The current implementation assumes the image / alpha / trimap
         a numpy array with pixel values ranging from 0 to 255.

From 7963b49bc553da66e8e1a65779258bfcbdf34093 Mon Sep 17 00:00:00 2001
From: xuan07472 <xuan07472@gmail.com>
Date: Fri, 13 Jan 2023 15:06:16 +0800
Subject: [PATCH 04/16] use try_import

---
 mmeval/metrics/gradient_error.py | 24 +++++++++++++++---------
 1 file changed, 15 insertions(+), 9 deletions(-)

diff --git a/mmeval/metrics/gradient_error.py b/mmeval/metrics/gradient_error.py
index 6cbe8659..1aa21a10 100644
--- a/mmeval/metrics/gradient_error.py
+++ b/mmeval/metrics/gradient_error.py
@@ -1,16 +1,21 @@
-# Copyright (c) OpenMMLab. All rights reserved.
-import cv2
+# Copyright (c) OpenMMLab. All rights reserved.Dict
 import numpy as np
-from typing import Dict, List, Sequence
+from typing import TYPE_CHECKING, Dict, List, Sequence
 
 from mmeval.core import BaseMetric
+from mmeval.utils import try_import
+
+if TYPE_CHECKING:
+    import cv2
+else:
+    cv2 = try_import('cv2')
 
 
 def gaussian(x: np.ndarray, sigma: float):
     """Gaussian function.
 
     Args:
-        x (array_like): The independent variable.
+        x (np.ndarray): The independent variable.
         sigma (float): Standard deviation of the gaussian function.
 
     Returns:
@@ -24,7 +29,7 @@ def dgaussian(x: np.ndarray, sigma: float):
     """Derivative of Gaussian function.
 
     Args:
-        x (array_like): The independent variable.
+        x (np.ndarray): The independent variable.
         sigma (float): Standard deviation of the gaussian function.
 
     Returns:
@@ -66,10 +71,7 @@ def gauss_filter(sigma: float, epsilon=1e-2):
 
 
 def gauss_gradient(img: np.ndarray, sigma: float):
-    """Gaussian gradient.The first order Gaussian derivative convolution
-    calculation is carried out by using Gaussian filter. Calculate their
-    gradients separately, make a difference, and then accumulate their squares.
-    The more similar the two, the smaller the Gradient error.
+    """Gaussian gradient.
 
     Reference: https://www.mathworks.com/matlabcentral/mlc-downloads/downloads/
     submissions/8060/versions/2/previews/gaussgradient/gaussgradient.m/
@@ -135,6 +137,10 @@ def __init__(self,
         self.sigma = sigma
         self.norm_const = norm_const
 
+        if cv2 is None:
+            raise ImportError(f'For availability of {self.__class__.__name__},'
+                              ' please install cv2 first.')
+
     def add(self, pred_alphas: Sequence[np.ndarray], gt_alphas: Sequence[np.ndarray], trimaps: Sequence[np.ndarray]) -> None:  # type: ignore # yapf: disable # noqa: E501
         """Add GradientError score of batch to ``self._results``
 

From 4503e53e315000416465046ee858950cd3a1d266 Mon Sep 17 00:00:00 2001
From: TKAN <40657922+xuan07472@users.noreply.github.com>
Date: Fri, 13 Jan 2023 15:23:29 +0800
Subject: [PATCH 05/16] Update gradient_error.py

update
---
 mmeval/metrics/gradient_error.py | 1 +
 1 file changed, 1 insertion(+)

diff --git a/mmeval/metrics/gradient_error.py b/mmeval/metrics/gradient_error.py
index 1aa21a10..a90a6fb4 100644
--- a/mmeval/metrics/gradient_error.py
+++ b/mmeval/metrics/gradient_error.py
@@ -116,6 +116,7 @@ class GradientError(BaseMetric):
         unknown and the background respectively.
 
     Examples:
+    
         >>> from mmeval import GradientError
         >>> import numpy as np
         >>>

From 651168abc72ef6a145d35db76383adbf5cdcc8fd Mon Sep 17 00:00:00 2001
From: xuan07472 <xuan07472@gmail.com>
Date: Fri, 13 Jan 2023 15:40:38 +0800
Subject: [PATCH 06/16] update

---
 mmeval/metrics/gradient_error.py | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/mmeval/metrics/gradient_error.py b/mmeval/metrics/gradient_error.py
index a90a6fb4..d34f1395 100644
--- a/mmeval/metrics/gradient_error.py
+++ b/mmeval/metrics/gradient_error.py
@@ -116,7 +116,7 @@ class GradientError(BaseMetric):
         unknown and the background respectively.
 
     Examples:
-    
+
         >>> from mmeval import GradientError
         >>> import numpy as np
         >>>

From afed4942778ac5345b5ae6ae5fd33cc3384a1a20 Mon Sep 17 00:00:00 2001
From: xuan07472 <xuan07472@gmail.com>
Date: Fri, 13 Jan 2023 15:47:49 +0800
Subject: [PATCH 07/16] update Examples

---
 mmeval/metrics/gradient_error.py | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

diff --git a/mmeval/metrics/gradient_error.py b/mmeval/metrics/gradient_error.py
index d34f1395..3f33769f 100644
--- a/mmeval/metrics/gradient_error.py
+++ b/mmeval/metrics/gradient_error.py
@@ -121,7 +121,8 @@ class GradientError(BaseMetric):
         >>> import numpy as np
         >>>
         >>> gradient_error = GradientError()
-        >>> pred_alpha = np.zeros((32, 32), dtype=np.uint8)
+        >>> np.random.seed(0)
+        >>> pred_alpha = np.random.randn(32, 32).astype('uint8')
         >>> gt_alpha = np.ones((32, 32), dtype=np.uint8) * 255
         >>> trimap = np.zeros((32, 32), dtype=np.uint8)
         >>> trimap[:16, :16] = 128

From 8c6e46ef44727e5c705115699b3961e32fcc899f Mon Sep 17 00:00:00 2001
From: xuan07472 <xuan07472@gmail.com>
Date: Fri, 13 Jan 2023 16:32:44 +0800
Subject: [PATCH 08/16] update test

---
 tests/test_metrics/test_gradient_error.py | 9 ++++-----
 1 file changed, 4 insertions(+), 5 deletions(-)

diff --git a/tests/test_metrics/test_gradient_error.py b/tests/test_metrics/test_gradient_error.py
index ab833c97..92dbe44d 100644
--- a/tests/test_metrics/test_gradient_error.py
+++ b/tests/test_metrics/test_gradient_error.py
@@ -12,8 +12,7 @@ def test_gradient_error():
     trimap[:16, :16] = 128
     trimap[16:, 16:] = 255
 
-    gradienterror = GradientError()
-    gradienterror_results = gradienterror(pred_alpha, gt_alpha, trimap)
-    assert isinstance(gradienterror_results, dict)
-    np.testing.assert_almost_equal(gradienterror_results['GradientError'],
-                                   0.0935)
+    gradient_error = GradientError()
+    metric_results = gradient_error(pred_alpha, gt_alpha, trimap)
+    assert isinstance(metric_results, dict)
+    np.testing.assert_almost_equal(metric_results['GradientError'], 0.0935)

From 086c77943aa6e4adca356274416a5729f7e7e11f Mon Sep 17 00:00:00 2001
From: TKAN <40657922+xuan07472@users.noreply.github.com>
Date: Fri, 13 Jan 2023 18:13:18 +0800
Subject: [PATCH 09/16] Apply suggestions from code review

update

Co-authored-by: yancong <32220263+ice-tong@users.noreply.github.com>
---
 mmeval/metrics/gradient_error.py | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/mmeval/metrics/gradient_error.py b/mmeval/metrics/gradient_error.py
index 3f33769f..18b42c2a 100644
--- a/mmeval/metrics/gradient_error.py
+++ b/mmeval/metrics/gradient_error.py
@@ -45,7 +45,7 @@ def gauss_filter(sigma: float, epsilon=1e-2):
     Args:
         sigma (float): Standard deviation of the gaussian kernel.
         epsilon (float): Small value used when calculating kernel size.
-            Default: 1e-2.
+            Default to 1e-2.
 
     Returns:
         tuple(np.ndarray, np.ndarray): Gaussian filter along x and y axis.
@@ -100,7 +100,7 @@ class GradientError(BaseMetric):
         sigma (float): Standard deviation of the gaussian kernel.
             Defaults to 1.4 .
         norm_const (int): Divide the result to reduce its magnitude.
-            Defaults to 1000 .
+            Defaults to 1000.
         **kwargs: Keyword parameters passed to :class:`BaseMetric`.
 
     Note:

From 412d91f0abacfb39c76797e4166e93d29a63bd79 Mon Sep 17 00:00:00 2001
From: TKAN <40657922+xuan07472@users.noreply.github.com>
Date: Fri, 13 Jan 2023 21:00:37 +0800
Subject: [PATCH 10/16] Update mmeval/metrics/gradient_error.py

Co-authored-by: yancong <32220263+ice-tong@users.noreply.github.com>
---
 mmeval/metrics/gradient_error.py | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/mmeval/metrics/gradient_error.py b/mmeval/metrics/gradient_error.py
index 18b42c2a..3e61c2e6 100644
--- a/mmeval/metrics/gradient_error.py
+++ b/mmeval/metrics/gradient_error.py
@@ -39,7 +39,7 @@ def dgaussian(x: np.ndarray, sigma: float):
     return -x * gaussian(x, sigma) / sigma**2
 
 
-def gauss_filter(sigma: float, epsilon=1e-2):
+def gauss_filter(sigma: float, epsilon: float = 1e-2):
     """Gaussian Filter.
 
     Args:

From 6e48201ff858ab79ef97325905ccebe9f7908d9c Mon Sep 17 00:00:00 2001
From: xuan07472 <xuan07472@gmail.com>
Date: Fri, 13 Jan 2023 21:38:57 +0800
Subject: [PATCH 11/16] Correct the Args

---
 mmeval/metrics/gradient_error.py | 11 +++++++----
 1 file changed, 7 insertions(+), 4 deletions(-)

diff --git a/mmeval/metrics/gradient_error.py b/mmeval/metrics/gradient_error.py
index 3e61c2e6..041447d1 100644
--- a/mmeval/metrics/gradient_error.py
+++ b/mmeval/metrics/gradient_error.py
@@ -147,9 +147,12 @@ def add(self, pred_alphas: Sequence[np.ndarray], gt_alphas: Sequence[np.ndarray]
         """Add GradientError score of batch to ``self._results``
 
         Args:
-            pred_alphas (Sequence[np.ndarray]): Pred_alpha data of predictions.
-            ori_alphas (Sequence[np.ndarray]): Ori_alpha data of data_batch.
-            ori_trimaps (Sequence[np.ndarray]): Ori_trimap data of data_batch.
+            pred_alphas (Sequence[np.ndarray]): Predict the probability
+                that pixels belong to the foreground.
+            gt_alphas (Sequence[np.ndarray]): Probability that the actual
+                pixel belongs to the foreground.
+            trimaps (Sequence[np.ndarray]): Broadly speaking, the trimap
+                consists of foreground and unknown region.
         """
 
         for pred_alpha, gt_alpha, trimap in zip(pred_alphas, gt_alphas,
@@ -189,4 +192,4 @@ def compute_metric(self, results: List) -> Dict[str, float]:
             and the values are corresponding results.
         """
 
-        return {'GradientError': float(np.array(results).mean())}
+        return {'gradient_error': float(np.array(results).mean())}

From 60dbe9a580500345e71ada2cebaab1420f387ac7 Mon Sep 17 00:00:00 2001
From: xuan07472 <xuan07472@gmail.com>
Date: Fri, 13 Jan 2023 23:50:36 +0800
Subject: [PATCH 12/16] update test

---
 tests/test_metrics/test_gradient_error.py | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/tests/test_metrics/test_gradient_error.py b/tests/test_metrics/test_gradient_error.py
index 92dbe44d..e6e9bfe0 100644
--- a/tests/test_metrics/test_gradient_error.py
+++ b/tests/test_metrics/test_gradient_error.py
@@ -15,4 +15,4 @@ def test_gradient_error():
     gradient_error = GradientError()
     metric_results = gradient_error(pred_alpha, gt_alpha, trimap)
     assert isinstance(metric_results, dict)
-    np.testing.assert_almost_equal(metric_results['GradientError'], 0.0935)
+    np.testing.assert_almost_equal(metric_results['gradient_error'], 0.0935)

From 016c5eb0459e34a1ec2221a4ba09d1b2c569a85b Mon Sep 17 00:00:00 2001
From: xuan07472 <xuan07472@gmail.com>
Date: Sat, 14 Jan 2023 17:13:39 +0800
Subject: [PATCH 13/16] Change metrics to lowercase

---
 mmeval/metrics/gradient_error.py | 10 +++++-----
 1 file changed, 5 insertions(+), 5 deletions(-)

diff --git a/mmeval/metrics/gradient_error.py b/mmeval/metrics/gradient_error.py
index 041447d1..704532b6 100644
--- a/mmeval/metrics/gradient_error.py
+++ b/mmeval/metrics/gradient_error.py
@@ -128,7 +128,7 @@ class GradientError(BaseMetric):
         >>> trimap[:16, :16] = 128
         >>> trimap[16:, 16:] = 255
         >>> gradient_error(pred_alpha, gt_alpha, trimap)  # doctest: +ELLIPSIS
-        {'GradientError': ...}
+        {'gradient_error': ...}
     """
 
     def __init__(self,
@@ -144,7 +144,7 @@ def __init__(self,
                               ' please install cv2 first.')
 
     def add(self, pred_alphas: Sequence[np.ndarray], gt_alphas: Sequence[np.ndarray], trimaps: Sequence[np.ndarray]) -> None:  # type: ignore # yapf: disable # noqa: E501
-        """Add GradientError score of batch to ``self._results``
+        """Add gradient_error score of batch to ``self._results``
 
         Args:
             pred_alphas (Sequence[np.ndarray]): Predict the probability
@@ -180,14 +180,14 @@ def add(self, pred_alphas: Sequence[np.ndarray], gt_alphas: Sequence[np.ndarray]
             self._results.append(grad_loss)
 
     def compute_metric(self, results: List) -> Dict[str, float]:
-        """Compute the GradientError metric.
+        """Compute the gradient_error metric.
 
         Args:
-            results (List): A list that consisting the GradientError score.
+            results (List): A list that consisting the gradient_error score.
                 This list has already been synced across all ranks.
 
         Returns:
-            Dict[str, float]: The computed GradientError metric.
+            Dict[str, float]: The computed gradient_error metric.
             The keys are the names of the metrics,
             and the values are corresponding results.
         """

From 23d4014b521d86010d308f0e0bc3737afdcab279 Mon Sep 17 00:00:00 2001
From: xuan07472 <xuan07472@gmail.com>
Date: Sat, 14 Jan 2023 18:38:14 +0800
Subject: [PATCH 14/16] update

---
 mmeval/metrics/gradient_error.py | 8 ++++----
 1 file changed, 4 insertions(+), 4 deletions(-)

diff --git a/mmeval/metrics/gradient_error.py b/mmeval/metrics/gradient_error.py
index 704532b6..e5376f4e 100644
--- a/mmeval/metrics/gradient_error.py
+++ b/mmeval/metrics/gradient_error.py
@@ -144,7 +144,7 @@ def __init__(self,
                               ' please install cv2 first.')
 
     def add(self, pred_alphas: Sequence[np.ndarray], gt_alphas: Sequence[np.ndarray], trimaps: Sequence[np.ndarray]) -> None:  # type: ignore # yapf: disable # noqa: E501
-        """Add gradient_error score of batch to ``self._results``
+        """Add GradientError score of batch to ``self._results``
 
         Args:
             pred_alphas (Sequence[np.ndarray]): Predict the probability
@@ -180,14 +180,14 @@ def add(self, pred_alphas: Sequence[np.ndarray], gt_alphas: Sequence[np.ndarray]
             self._results.append(grad_loss)
 
     def compute_metric(self, results: List) -> Dict[str, float]:
-        """Compute the gradient_error metric.
+        """Compute the GradientError metric.
 
         Args:
-            results (List): A list that consisting the gradient_error score.
+            results (List): A list that consisting the GradientError score.
                 This list has already been synced across all ranks.
 
         Returns:
-            Dict[str, float]: The computed gradient_error metric.
+            Dict[str, float]: The computed GradientError metric.
             The keys are the names of the metrics,
             and the values are corresponding results.
         """

From 44ed2578d217b6453babc935e92448868475370b Mon Sep 17 00:00:00 2001
From: Zaida Zhou <58739961+zhouzaida@users.noreply.github.com>
Date: Sun, 15 Jan 2023 11:36:14 +0800
Subject: [PATCH 15/16] Update mmeval/metrics/gradient_error.py

---
 mmeval/metrics/gradient_error.py | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/mmeval/metrics/gradient_error.py b/mmeval/metrics/gradient_error.py
index e5376f4e..99b559fc 100644
--- a/mmeval/metrics/gradient_error.py
+++ b/mmeval/metrics/gradient_error.py
@@ -141,7 +141,7 @@ def __init__(self,
 
         if cv2 is None:
             raise ImportError(f'For availability of {self.__class__.__name__},'
-                              ' please install cv2 first.')
+                              ' please install opencv-python first.')
 
     def add(self, pred_alphas: Sequence[np.ndarray], gt_alphas: Sequence[np.ndarray], trimaps: Sequence[np.ndarray]) -> None:  # type: ignore # yapf: disable # noqa: E501
         """Add GradientError score of batch to ``self._results``

From 9267768ce286c650e548d014b3d4c6a74171c441 Mon Sep 17 00:00:00 2001
From: Zaida Zhou <58739961+zhouzaida@users.noreply.github.com>
Date: Sun, 15 Jan 2023 11:48:23 +0800
Subject: [PATCH 16/16] Apply suggestions from code review

---
 docs/en/api/metrics.rst    | 2 +-
 docs/zh_cn/api/metrics.rst | 2 +-
 mmeval/metrics/__init__.py | 2 +-
 3 files changed, 3 insertions(+), 3 deletions(-)

diff --git a/docs/en/api/metrics.rst b/docs/en/api/metrics.rst
index 583b8a0d..00d24b3b 100644
--- a/docs/en/api/metrics.rst
+++ b/docs/en/api/metrics.rst
@@ -43,4 +43,4 @@ Metrics
    MSE
    BLEU
    SAD
-   GradientError
\ No newline at end of file
+   GradientError
diff --git a/docs/zh_cn/api/metrics.rst b/docs/zh_cn/api/metrics.rst
index 583b8a0d..00d24b3b 100644
--- a/docs/zh_cn/api/metrics.rst
+++ b/docs/zh_cn/api/metrics.rst
@@ -43,4 +43,4 @@ Metrics
    MSE
    BLEU
    SAD
-   GradientError
\ No newline at end of file
+   GradientError
diff --git a/mmeval/metrics/__init__.py b/mmeval/metrics/__init__.py
index f94eb4c3..2479d788 100644
--- a/mmeval/metrics/__init__.py
+++ b/mmeval/metrics/__init__.py
@@ -27,5 +27,5 @@
     'F1Metric', 'HmeanIoU', 'SingleLabelMetric', 'COCODetectionMetric',
     'PCKAccuracy', 'MpiiPCKAccuracy', 'JhmdbPCKAccuracy', 'ProposalRecall',
     'PSNR', 'MAE', 'MSE', 'SSIM', 'SNR', 'MultiLabelMetric',
-    'AveragePrecision', 'AVAMeanAP', 'BLEU', 'SAD', 'GradientError
+    'AveragePrecision', 'AVAMeanAP', 'BLEU', 'SAD', 'GradientError'
 ]