python - 值错误输入 0 与层模型不兼容

Question

我正在尝试使用深度学习进行肺部检测，但遇到了一些错误。

我有一个 keras 模型，我将附在下面。我得到的错误如下：

ValueError: Input 0 is incompatible with layer model: expected shape=(None, 1, 512, 512), found shape=(2, 3, 512, 512)

我拥有的模型如下：

def get_unet():
inputs = Input((1,img_rows, img_cols))
conv1 = Convolution2D(32, 3, 1, activation='relu', padding='same')(inputs)
conv1 = Convolution2D(32, 3, 1, activation='relu', padding='same')(conv1)
pool1 = MaxPooling2D(pool_size=(2, 2))(conv1)

conv2 = Convolution2D(64, 3, 1, activation='relu', padding='same')(pool1)
conv2 = Convolution2D(64, 3, 1, activation='relu', padding='same')(conv2)
pool2 = MaxPooling2D(pool_size=(2, 2))(conv2)

conv3 = Convolution2D(128, 3, 1, activation='relu', padding='same')(pool2)
conv3 = Convolution2D(128, 3, 1, activation='relu', padding='same')(conv3)
pool3 = MaxPooling2D(pool_size=(2, 2), padding='same')(conv3)

conv4 = Convolution2D(256, 3, 1, activation='relu', padding='same')(pool3)
conv4 = Convolution2D(256, 3, 1, activation='relu', padding='same')(conv4)
pool4 = MaxPooling2D(pool_size=(2, 2), padding='same')(conv4)

conv5 = Convolution2D(512, 3, 1, activation='relu', padding='same')(pool4)
conv5 = Convolution2D(512, 3, 1, activation='relu', padding='same')(conv5)
up6 = concatenate([UpSampling2D(size=(2, 2))(conv5), conv4], axis=1)

conv6 = Convolution2D(256, 3, 1, activation='relu', padding='same')(up6)
conv6 = Convolution2D(256, 3, 1, activation='relu', padding='same')(conv6)

up7 = concatenate([UpSampling2D(size=(2, 2))(conv6), conv3], axis=1)
conv7 = Convolution2D(128, 3, 1, activation='relu', padding='same')(up7)
conv7 = Convolution2D(128, 3, 1, activation='relu', padding='same')(conv7)

up8 = concatenate([UpSampling2D(size=(2, 2))(conv7), conv2], axis=1)
conv8 = Convolution2D(64, 3, 1, activation='relu', padding='same')(up8)
conv8 = Convolution2D(64, 3, 1, activation='relu', padding='same')(conv8)

up9 = concatenate([UpSampling2D(size=(2, 2))(conv8), conv1], axis=1)
conv9 = Convolution2D(32, 3, 1, activation='relu', padding='same')(up9)
conv9 = Convolution2D(32, 3, 1, activation='relu', padding='same')(conv9)

conv10 = Convolution2D(1, 1, 1, activation='sigmoid')(conv9)

model = Model(inputs=inputs, outputs=conv10)
print(model.summary())

model.compile(optimizer=Adam(lr=1.0e-5), loss=dice_coef_loss, metrics=[dice_coef])

return model

我知道模型将输入作为（1,512,512），图像的输入是（3,512,512）。我只是一个初学者，我不知道该怎么做。但是，我不想更改模型的结构并希望保持原样。我也不想改变模型输入的形状。所以请给我一些改变图像尺寸的建议，使它们与模型的输入尺寸相匹配！谢谢

标签： pythondeep-learningneural-networkconv-neural-network