def __call__(self, img: np.ndarray):
"""
对图片进行处理，先按照高度进行resize，resize之后如果宽度不足指定宽度，就补黑色像素，否则就强行缩放到指定宽度
:param img_path: 图片地址
:return:
"""
data_augment = False
if self.phase == 'train' and np.random.rand() > 0.5:
data_augment = True
if data_augment:
img_h = 40
img_w = 340
else:
img_h = self.img_h
img_w = self.img_w
h, w = img.shape[:2]
ratio_h = float(img_h) / h
new_w = int(w * ratio_h)
if new_w < img_w and self.pad:
img = cv2.resize(img, (new_w, img_h))
if len(img.shape) == 2:
img = np.expand_dims(img, 3)
step = np.zeros((img_h, img_w - new_w, img.shape[-1]), dtype=img.dtype)
img = np.column_stack((img, step))
else:
img = cv2.resize(img, (img_w, img_h))
if data_augment:
img = nd.array(img)
img, _ = image.random_crop(img, (self.img_w, self.img_h))
img = img.asnumpy()
return img
def _empirical_cdf_inverse_transform(
batch_target_sorted: np.ndarray,
batch_predictions: np.ndarray,
slopes: np.ndarray,
intercepts: np.ndarray,
) -> np.ndarray:
"""
Apply forward transformation of the empirical CDF.
Parameters
----------
batch_target_sorted
Sorted targets of the input batch.
batch_predictions
Predictions of the underlying probability distribution
slopes
Slopes of the piece-wise linear function.
intercepts
Intercepts of the piece-wise linear function.
Returns
-------
outputs
Forward transformed outputs.
"""
slopes = slopes.asnumpy()
intercepts = intercepts.asnumpy()
batch_target_sorted = batch_target_sorted.asnumpy()
batch_size, num_timesteps, target_dim = batch_target_sorted.shape
indices = np.floor(batch_predictions * num_timesteps)
# indices = indices - 1
# for now project into [0, 1]
indices = np.clip(indices, 0, num_timesteps - 1)
indices = indices.astype(np.int)
transformed = np.where(
np.take_along_axis(slopes, indices, axis=1) != 0.0,
(batch_predictions -
np.take_along_axis(intercepts, indices, axis=1)) /
np.take_along_axis(slopes, indices, axis=1),
np.take_along_axis(batch_target_sorted, indices, axis=1),
)
return transformed
