我最近了解到焦点损失函数，听说它主要用于处理不平衡数据集。所以我尝试在Cifar10数据集上使用我在网上找到的这个简单的焦点损失函数（适用于Keras）。
我不断遇到一个错误，错误信息我已经在最后提到了。我尝试了几种方法来解决它，但都没有成功。请帮我看一下，我非常感谢你的帮助。谢谢你！
焦点损失
输入数据
from keras.datasets import cifar10(xtrain,ytrain),(xtest,ytest) = cifar10.load_data()神经网络
from keras.layers import Dense, Conv2D, Flatten, MaxPool2Dfrom keras.models import Sequentialfrom keras.optimizers import Adammodel = Sequential([ Conv2D(filters=64, kernel_size=(27,27), strides=(1,1), input_shape=(32,32,3),padding='same', activation='sigmoid'), MaxPool2D(pool_size=(13,13), strides=(1,1), padding='valid'), Conv2D(filters=32, kernel_size=(11,11), strides=(1,1), padding='valid', activation='sigmoid'), Flatten(), Dense(units=600, activation='sigmoid'), Dense(units=128, activation='sigmoid'), Dense(units=10, activation='softmax')])编译和拟合
model.compile(loss=FocalLoss, optimizer=Adam(learning_rate=0.0001), metrics=['accuracy'])model.fit(xtrain, ytrain, epochs=10, batch_size=120, validation_data=(xtest,ytest), verbose=2)拟合时出现的错误
Epoch 1/10---------------------------------------------------------------------------TypeError Traceback (most recent call last)<ipython-input-52-52246069690d> in <module>()----> 1 model.fit(xtrain, ytrain, epochs=10, batch_size=120, validation_data=(xtest,ytest), verbose=2)10 frames/usr/local/lib/python3.6/dist-packages/tensorflow/python/framework/func_graph.py in wrapper(*args, **kwargs) 971 except Exception as e: # pylint:disable=broad-except 972 if hasattr(e, "ag_error_metadata"):--> 973 raise e.ag_error_metadata.to_exception(e) 974 else: 975 raiseTypeError: in user code: /usr/local/lib/python3.6/dist-packages/tensorflow/python/keras/engine/training.py:806 train_function * return step_function(self, iterator) <ipython-input-50-e8cbeb45fe58>:12 FocalLoss * BCE = K.binary_crossentropy(targets, inputs) /usr/local/lib/python3.6/dist-packages/tensorflow/python/util/dispatch.py:201 wrapper ** return target(*args, **kwargs) /usr/local/lib/python3.6/dist-packages/tensorflow/python/keras/backend.py:4829 binary_crossentropy bce = target * math_ops.log(output + epsilon()) /usr/local/lib/python3.6/dist-packages/tensorflow/python/ops/math_ops.py:1141 binary_op_wrapper raise e /usr/local/lib/python3.6/dist-packages/tensorflow/python/ops/math_ops.py:1125 binary_op_wrapper return func(x, y, name=name) /usr/local/lib/python3.6/dist-packages/tensorflow/python/ops/math_ops.py:1457 _mul_dispatch return multiply(x, y, name=name) /usr/local/lib/python3.6/dist-packages/tensorflow/python/util/dispatch.py:201 wrapper return target(*args, **kwargs) /usr/local/lib/python3.6/dist-packages/tensorflow/python/ops/math_ops.py:509 multiply return gen_math_ops.mul(x, y, name=name) /usr/local/lib/python3.6/dist-packages/tensorflow/python/ops/gen_math_ops.py:6176 mul "Mul", x=x, y=y, name=name) /usr/local/lib/python3.6/dist-packages/tensorflow/python/framework/op_def_library.py:506 _apply_op_helper inferred_from[input_arg.type_attr])) TypeError: Input 'y' of 'Mul' Op has type float32 that does not match type uint8 of argument 'x'.注意
xtrain和ytrain的dtype是相同的，即’uint8′
回答：
问题与你的目标类型有关，它们是int8类型，但你需要将它们转换为float32类型。我在损失函数内部进行了转换，并且移除了flatten部分，因为那是错误的
def FocalLoss(targets, inputs, alpha=ALPHA, gamma=GAMMA): targets = K.cast(targets, 'float32') BCE = K.binary_crossentropy(targets, inputs) BCE_EXP = K.exp(-BCE) focal_loss = K.mean(alpha * K.pow((1-BCE_EXP), gamma) * BCE) return focal_loss这里是运行的笔记本：https://colab.research.google.com/drive/1E89tggfCvifuoJRdGuXTHuBQPvXFCYN4?usp=sharing
