如何将 fit_generator 与多个输入一起使用

是否可以有两个 fit_generator？

我正在创建一个有两个输入的模型， 模型配置如下所示。

标签 Y 对 X1 和 X2 数据使用相同的标签。

将会继续出现以下错误。

检查模型输入时出错：您传递给模型的 Numpy 数组列表不是模型预期的大小。预期的 查看 2 个数组，但得到以下 1 个数组的列表： [数组([[[0.75686276, 0.75686276, 0.75686276], [0.75686276, 0.75686276, 0.75686276], [0.75686276, 0.75686276, 0.75686276], ..., [0.65882355, 0.65882355, 0.65882355...

我的代码如下所示：

def generator_two_img(X1, X2, Y,batch_size):
    generator = ImageDataGenerator(rotation_range=15,
                                   width_shift_range=0.2,
                                   height_shift_range=0.2,
                                   shear_range=0.2,
                                   zoom_range=0.2,
                                   horizontal_flip=True,
                                   fill_mode='nearest')

    genX1 = generator.flow(X1, Y, batch_size=batch_size)
    genX2 = generator.flow(X2, Y, batch_size=batch_size)

    while True:
        X1 = genX1.__next__()
        X2 = genX2.__next__()
        yield [X1, X2], Y
  """
      .................................
  """
hist = model.fit_generator(generator_two_img(x_train, x_train_landmark, 
                y_train, batch_size),
                steps_per_epoch=len(x_train) // batch_size, epochs=nb_epoch,
                callbacks = callbacks,
                validation_data=(x_validation, y_validation),
                validation_steps=x_validation.shape[0] // batch_size, 
                `enter code here`verbose=1)

试试这个生成器：

def generator_two_img(X1, X2, y, batch_size):
    genX1 = gen.flow(X1, y,  batch_size=batch_size, seed=1)
    genX2 = gen.flow(X2, y, batch_size=batch_size, seed=1)
    while True:
        X1i = genX1.next()
        X2i = genX2.next()
        yield [X1i[0], X2i[0]], X1i[1]

3 个输入的发生器：

def generator_three_img(X1, X2, X3, y, batch_size):
    genX1 = gen.flow(X1, y,  batch_size=batch_size, seed=1)
    genX2 = gen.flow(X2, y, batch_size=batch_size, seed=1)
    genX3 = gen.flow(X3, y, batch_size=batch_size, seed=1)
    while True:
        X1i = genX1.next()
        X2i = genX2.next()
        X3i = genX3.next()
        yield [X1i[0], X2i[0], X3i[0]], X1i[1]

EDIT（添加生成器、输出图像和 numpy 数组以及目标）

#X1 is an image, y is the target, X2 is a numpy array - other data input        
def gen_flow_for_two_inputs(X1, X2, y):
    genX1 = gen.flow(X1,y,  batch_size=batch_size,seed=666)
    genX2 = gen.flow(X1,X2, batch_size=batch_size,seed=666)
    while True:
        X1i = genX1.next()
        X2i = genX2.next()
        #Assert arrasy are equal - this was for peace of mind, but slows down training
        #np.testing.assert_array_equal(X1i[0],X2i[0])
        yield [X1i[0], X2i[1]], X1i[1]