NotImplementedError:无法将符号张量(up_sampling2d_4_target:0)转换为numpy数组

Question

NotImplementedError:无法将符号张量(up_sampling2d_4_target:0)转换为numpy数组。

跟踪误差

import keras.backend as K
from keras.optimizers import Adam
from keras.losses import binary_crossentropy

## intersection over union
def IoU(y_true, y_pred, eps=1e-6):
    if np.max(y_true) == 0.0:
        return IoU(1-y_true, 1-y_pred) ## empty image; calc IoU of zeros
    intersection = K.sum(y_true * y_pred, axis=[1,2,3])
    union = K.sum(y_true, axis=[1,2,3]) + K.sum(y_pred, axis=[1,2,3]) - intersection
    return -K.mean( (intersection + eps) / (union + eps), axis=0)

NotImplementedError跟踪(最近一次调用)在1415年，而True：--> 16loss_history = fit() 17如果np.min([mh.History‘’val_loss_history中的mh ])< -0.2: 18中断

在fit() 1 def fit()中：->2 seg_model.compile(optimizer=Adam(1e-3，衰变=1e-6)，loss=IoU，度量=‘二进制_精度’)34 step_count = min(MAX_TRAIN_STEPS，train_df.form//批_SIZE)5 aug_gen = create_aug_gen(make_image_gen(train_df))

~/venv/lib/python3.7/site-packages/tensorflow_core/python/training/tracking/base.py in _method_wrapper(self，*args，**kwargs) 455 self._self_setattr_tracking = False # pylint: self._self_setattr_tracking=Protected-Access456 self._self_setattr_tracking：-> 457 self._self_setattr_tracking= method(self，*args，**kwargs) 458最后: 459 self._self_setattr_tracking = previous_value # pylint: disable=protected-access

编译中的~/venv/lib/python3.7/site-packages/tensorflow_core/python/keras/engine/training.py (自、优化器、丢失、度量、loss_weights、sample_weight_mode、weighted_metrics、target_tensors、distribute、**kwargs) 371 372 #创建了模型损失和加权度量子图。-> 373 self._compile_weights_loss_and_weighted_metrics() 374 375 #火车、测试和预测意志函数

~/venv/lib/python3.7/site-packages/tensorflow_core/python/training/tracking/base.py in _method_wrapper(self，*args，**kwargs) 455 self._self_setattr_tracking = False # pylint: self._self_setattr_tracking=Protected-Access456 self._self_setattr_tracking：-> 457 self._self_setattr_tracking= method(self，*args，**kwargs) 458最后: 459 self._self_setattr_tracking = previous_value # pylint: disable=protected-access

~/venv/lib/python3.7/site-packages/tensorflow_core/python/keras/engine/training.py in _compile_weights_loss_and_weighted_metrics(self，sample_weights)

1651 # loss_weight_2 *output_2_loss_fn(.)+ 1652 #层损失-> 1653 self.total_loss =self._prepare_total_loss(掩码) 1654 1655 def _prepare_skip_target_masks(self)：

~/venv/lib/python3.7/site-packages/tensorflow_core/python/keras/engine/training.py in _prepare_total_loss(自我，面具)17111712(loss_fn，'reduction')：-> 1713 per_sample_losses = loss_fn.call(y_truey_pred) 1714 weighted_losses = losses_utils.compute_weighted_loss( 1715 per_sample_losses，

~/venv/lib/python3.7/site-packages/tensorflow_core/python/keras/losses.py in call(self，y_true，y_pred) 219 y_pred，y_true = tf_losses_util.squeeze_or_expand_dimensions( 220 y_pred，y_true) -> 221返回self.fn(y_true，y_pred，**self._fn_kwargs) 22223def get_config(self)：

在IoU(y_true，y_pred，eps) 5 ##交集中，在## 6 def IoU(y_true，y_pred，eps=1e-6)中：-如果np.max(y_true) == 0.0: 8返回IoU(1-y_true，1-y_pred) ##空图像；calc IoU of zeros 9交叉口= K.sum(y_true * y_pred，axis=1,2,3)

<array_function internals> in amax(*args，**kwargs)

在amax中的~/venv/lib/python3.7/site-packages/numpy/core/fromnumeric.py (a，轴，输出，保持，初始值，其中) 2619“2620

返回_wrapreduction(a，np.maximum，'max'，轴，无，out，-> 2621 keepdims=keepdims，initial=initial，where=where) 2622 2623

~/venv/lib/python3.7/site-packages/numpy/core/fromnumeric.py in _wrapreduction(obj，ufunc，方法，轴，dtype，out，**kwargs) 88返回减少(axis=axis，out=out，**passkwargs) 89 -> 90返回ufunc.reduce(obj，axis，dtype，out，**passkwargs) 91 92

~/venv/lib/python3.7/site-packages/tensorflow_core/python/framework/ops.py in (Self) 734 def array(self)：735引发NotImplementedError(“无法将符号张量({})转换为numpy”数组--> 736“数组”。“.format(self.name)”737 738 def len(self)：

NotImplementedError:无法将符号张量(up_sampling2d_4_target:0)转换为numpy

阵列。

Answer 1

不能将numpy与tensorflow Tensor结合使用--它们是两种不同的东西。

问题在于：

if np.max(y_true) == 0.0:
    return IoU(1-y_true, 1-y_pred) ## empty image; calc IoU of zeros

相反，您需要这些行：

is_zero = K.equal(y_true, 0)
y_true = K.switch(is_zero, 1-y_true, y_true)
y_pred = K.switch(is_zero, 1-y_pred, y_pred)