如何为 Keras 模型编写基于二元交叉熵损失的条件回归损失函数

Question

我正在构建一个 key-point 人脸检测系统。目标是将面部图像输入模型，然后模型检测图像中的解剖标志（眼睛、鼻子）并输出可见标志的像素坐标。每个地标有三个目标：x、y、可见。 X 和 Y 是像素坐标，visible 是地标是否在图像中。该计划是首先在预测能见度和真实能见度之间进行二元交叉熵损失。然后，第二个损失是 x、y 坐标和目标之间的回归损失（我正在使用 MAPE）。但是，回归损失只会针对可见的地标进行计算。损失看起来像：

#Pseudo-code
def loss(y_true,y_pred):
    if y_true[2] == 1
       #Probability that landmark is in image
       #Compute binary cross entropy loss
       #Compute MAPE regression loss
       Total_loss = Binary_loss + MAPE_loss
       return Total_loss

    else:
       Total_loss = Binary loss
       return Total_loss

一旦编写了损失函数，我将如何在代码中实现它？我知道如何为每个问题创建模型（检查坐标，并分别检查可见性），但我不确定如何将两个头与条件损失函数结合起来。我将如何组合层（每个头的 Conv、Flatten、Dense）以获得所需的输出？谢谢！

编辑：我无法上传数据，但这是它的图像。前 9 列是地标的坐标和可见性。最后一列是已展平的相应图像。当我加载数据进行训练时，这些是我执行的步骤：

###Read in data file
file = "Directory/file.csv"
train_data = pd.read_csv(file)
###Convert each coordinate column to type float64
train_data['xreye'] = train_data['xreye'].astype(np.float64)
...
###Convert image column to string type
train_data['Image'] = train_data['Image'].astype(str)

#Image is feature, other values are labels to predict later
#Image column values are strings, also some missing values, have to split
##string by space and append it and handle missing values
imag = []
for i in range(len(train_data)):
    img = train_data['Image'][i].split(' ')
    img = ['0' if x == '' else x for x in img]      
    imag.append(img) 
#Reshape and convert to float value
image_list = np.array(imag,dtype = 'uint8')
X_train = image_list.reshape(-1,256,256,1)

####Get pixel coordinates and visibility targets
training = train_data[['xreye','yreye','reyev','xleye','yleye','leyev','xtsept','ytsept','tseptv']]
y_train = []
for i in range(len(train_data)):
    y = training.iloc[i,:]
    y_train.append(y)

y_train = np.array(y_train, dtype='float')

编辑：模型代码、损失函数和拟合方法。

###Loss function
visuals_mask = [False, False, True] * 3
def loss_func(y_true, y_pred):
    visuals_true = tf.boolean_mask(y_true, visuals_mask, axis=1)
    visuals_pred = tf.boolean_mask(y_pred, visuals_mask, axis=1)
    visuals_loss = tf.keras.losses.BinaryCrossentropy(visuals_true, visuals_pred)
    visuals_loss = tf.reduce_mean(visuals_loss)

    coords_true = tf.boolean_mask(y_true, ~np.array(visuals_mask), axis=1)
    coords_pred = tf.boolean_mask(y_pred, ~np.array(visuals_mask), axis=1)
    coords_loss = tf.keras.losses.MeanAbsolutePercentageError(coords_true, coords_pred)
    coords_loss = tf.reduce_mean(coords_loss)

    return coords_loss + visuals_loss
####Model code
model = Sequential()

model.add(Conv2D(32, (3,3), activation='relu', padding='same', use_bias=False, input_shape=(256,256,1)))
model.add(BatchNormalization())
model.add(MaxPool2D(pool_size=(2,2)))

model.add(Conv2D(64, (3,3), activation='relu', padding='same', use_bias=False))
model.add(BatchNormalization())
model.add(MaxPool2D(pool_size=(2,2)))

model.add(Conv2D(128, (3,3), activation='relu', padding='same', use_bias=False))
model.add(BatchNormalization())
model.add(MaxPool2D(pool_size=(2,2)))

model.add(Flatten())

model.add(Dense(128, activation='relu'))
model.add(Dense(64, activation='relu'))
model.add(Dense(32, activation='relu'))
model.add(Dense(16, activation='relu'))
model.add(Dense(8, activation='relu'))
model.add(Dense(4, activation='relu'))
model.add(Dense(9, activation='linear'))
model.summary()
model.compile(optimizer='adam', loss=loss_func)

###Model fit
checkpointer = ModelCheckpoint('C:/Users/Cloud/.spyder-py3/x_y_shift/weights/vis_coords_TEST.hdf5', monitor='val_loss', verbose=1, mode = 'min', save_best_only=True)
out = model.fit(X_train,y_train,epochs=5,batch_size=4,validation_split=0.1, verbose=1, callbacks=[checkpointer])

Answer 1

我不能确定，因为我没有重现问题的数据，但这些是我头脑中的步骤：

使用 boolean masking 从输出中获取 2、5 和 8 索引：

visuals_mask_ = [False, False, True] * 3

# in the loss function
visuals_true = tf.boolean_mask(y_true, visuals_mask_, axis=-1) # do the same with preds

计算视觉损失

visuals_loss = binary_crossentropy(visuals_true, visuals_pred) # use sparse if that's the case

获取坐标的输出，就像我们对视觉对象所做的那样，但使用相反的 visuals_mask。我相信 tf.boolean_mask(y_true, tf.math.logical_not(visuals_mask_, axis=-1)) 应该有效。
为其余部分（coords_true 和 coords_pred）计算 MAPE
通过 tf.reduce_mean
获取损失总和return它

我希望这些能提供一些见解。

编辑： 我尝试了以下方法，似乎有效：

y_true = tf.convert_to_tensor(np.random.rand(32, 9))
y_pred = tf.convert_to_tensor(np.random.rand(32, 9))

visuals_mask = [False, False, True] * 3

def loss_func(y_true, y_pred):
    visuals_true = tf.boolean_mask(y_true, visuals_mask, axis=1)
    visuals_pred = tf.boolean_mask(y_pred, visuals_mask, axis=1)
    visuals_loss = binary_crossentropy(visuals_true, visuals_pred)
    visuals_loss = tf.reduce_mean(visuals_loss)

    coords_true = tf.boolean_mask(y_true, ~np.array(visuals_mask), axis=1)
    coords_pred = tf.boolean_mask(y_pred, ~np.array(visuals_mask), axis=1)
    coords_loss = mean_absolute_percentage_error(coords_true, coords_pred)
    coords_loss = tf.reduce_mean(coords_loss)

    return coords_loss + visuals_loss

loss_func(y_true, y_pred)

我这里假设的是：

您的输出实际长度为 9 ((batch_size, 9))。
由于eager execution。

编辑 2： 我用这种模型试过它似乎有效:

model = Sequential()

model.add(Conv2D(4, 10, data_format='channels_last', input_shape=(256, 256, 1)))
model.add(Flatten())
model.add(Dense(9, activation='sigmoid'))

model.compile('adam', loss=loss_func)

如何为 Keras 模型编写基于二元交叉熵损失的条件回归损失函数

How to write a conditonal regressional loss function based on binary cross entropy loss for Keras model

python

keras

deep-learning

conv-neural-network

loss-function