python - Keras 中的 Segnet : total size of new array must be unchanged error

Question

我正在用 Python 实现 Segnet。以下是代码。

img_w = 480
img_h = 360
pool_size = 2

def build_model(img_w, img_h, pool_size):
    n_labels = 12

    kernel = 3

    encoding_layers = [
        Conv2D(64, (kernel, kernel), input_shape=(img_h, img_w, 3), padding='same'),
        BatchNormalization(),
        Activation('relu'),
        Convolution2D(64, (kernel, kernel), padding='same'),
        BatchNormalization(),
        Activation('relu'),
        MaxPooling2D(pool_size = (pool_size,pool_size)),

        Convolution2D(128, (kernel, kernel), padding='same'),
        BatchNormalization(),
        Activation('relu'),
        Convolution2D(128, (kernel, kernel), padding='same'),
        BatchNormalization(),
        Activation('relu'),
        MaxPooling2D(pool_size = (pool_size,pool_size)),

        Convolution2D(256, (kernel, kernel), padding='same'),
        BatchNormalization(),
        Activation('relu'),
        Convolution2D(256, (kernel, kernel), padding='same'),
        BatchNormalization(),
        Activation('relu'),
        Convolution2D(256, (kernel, kernel), padding='same'),
        BatchNormalization(),
        Activation('relu'),
        MaxPooling2D(pool_size = (pool_size,pool_size)),

        Convolution2D(512, (kernel, kernel), padding='same'),
        BatchNormalization(),
        Activation('relu'),
        Convolution2D(512, (kernel, kernel), padding='same'),
        BatchNormalization(),
        Activation('relu'),
        Convolution2D(512, (kernel, kernel), padding='same'),
        BatchNormalization(),
        Activation('relu'),
        MaxPooling2D(pool_size = (pool_size,pool_size)),

        Convolution2D(512, (kernel, kernel), padding='same'),
        BatchNormalization(),
        Activation('relu'),
        Convolution2D(512, (kernel, kernel), padding='same'),
        BatchNormalization(),
        Activation('relu'),
        Convolution2D(512, (kernel, kernel), padding='same'),
        BatchNormalization(),
        Activation('relu'),
        MaxPooling2D(pool_size = (pool_size,pool_size)),
    ]

    autoencoder = models.Sequential()
    autoencoder.encoding_layers = encoding_layers

    for l in autoencoder.encoding_layers:
        autoencoder.add(l)

    decoding_layers = [
        UpSampling2D(),
        Convolution2D(512, (kernel, kernel), padding='same'),
        BatchNormalization(),
        Activation('relu'),
        Convolution2D(512, (kernel, kernel), padding='same'),
        BatchNormalization(),
        Activation('relu'),
        Convolution2D(512, (kernel, kernel), padding='same'),
        BatchNormalization(),
        Activation('relu'),

        UpSampling2D(),
        Convolution2D(512, (kernel, kernel), padding='same'),
        BatchNormalization(),
        Activation('relu'),
        Convolution2D(512, (kernel, kernel), padding='same'),
        BatchNormalization(),
        Activation('relu'),
        Convolution2D(256, (kernel, kernel), padding='same'),
        BatchNormalization(),
        Activation('relu'),

        UpSampling2D(),
        Convolution2D(256, (kernel, kernel), padding='same'),
        BatchNormalization(),
        Activation('relu'),
        Convolution2D(256, (kernel, kernel), padding='same'),
        BatchNormalization(),
        Activation('relu'),
        Convolution2D(128, (kernel, kernel), padding='same'),
        BatchNormalization(),
        Activation('relu'),

        UpSampling2D(),
        Convolution2D(128, (kernel, kernel), padding='same'),
        BatchNormalization(),
        Activation('relu'),
        Convolution2D(64, (kernel, kernel), padding='same'),
        BatchNormalization(),
        Activation('relu'),

        UpSampling2D(),
        Convolution2D(64, (kernel, kernel), padding='same'),
        BatchNormalization(),
        Activation('relu'),
        Convolution2D(n_labels, (1, 1), padding='valid', activation="sigmoid"),
        BatchNormalization(),
    ]
    autoencoder.decoding_layers = decoding_layers
    for l in autoencoder.decoding_layers:
        autoencoder.add(l)

    autoencoder.add(Reshape((n_labels, img_h * img_w)))
    autoencoder.add(Permute((2, 1)))
    autoencoder.add(Activation('softmax'))



    return autoencoder

model = build_model(img_w, img_h, pool_size)

但它返回错误。

---------------------------------------------------------------------------
ValueError                                Traceback (most recent call last)
<ipython-input-21-051f06a53a14> in <module>()
----> 1 model = build_model(img_w, img_h, pool_size)

<ipython-input-20-c37fd94c8641> in build_model(img_w, img_h, pool_size)
    119         autoencoder.add(l)
    120 
--> 121     autoencoder.add(Reshape((n_labels, img_h * img_w)))
    122     autoencoder.add(Permute((2, 1)))
    123     autoencoder.add(Activation('softmax'))
    ValueError: total size of new array must be unchanged

我看不出错误的任何原因。当我将 img_w 和 img_h 更改为 256 时，此错误已解决，但问题是这不是图像大小或原始数据集，因此我无法使用它。如何解决？

Answer 1

问题是您正在执行 (2, 2) 下采样 5 次，所以让我们跟踪形状:

(360, 480) -> (180, 240) -> (90, 120) -> (45, 60) -> (22, 30) -> (11, 15)

现在上采样:

(11, 15) -> (22, 30) -> (44, 60) -> (88, 120) -> (176, 240) -> (352, 480)

因此，当您尝试使用原始形状reshape 输出时 - 由于模型不匹配而引发了问题。

可能的解决方案:

1. 调整图像大小，使两个输入维度都可以被 32 整除(例如 (352, 480) 或 (384, 480).

2. 在第 3 次上采样后添加 ZeroPadding2D(((1, 0), (0, 0))) 以更改 (44, 60) 的形状到 (45, 60)，什么将使您的网络以良好的输出形状结束。

其他问题:

请找出最后一个 MaxPooling2D 后面是第一个 Upsampling2D。这可能是个问题，因为这是您网络的无用瓶颈。