小编Sou*_*wal的帖子

我最近开始学习图像分割和UNet。我正在尝试进行多类图像分割，其中我有 7 个类，输入是 (256, 256, 3) rgb 图像，输出是 (256, 256, 1) 灰度图像，其中每个强度值对应于一个类。我正在做像素级的softmax。我使用稀疏分类交叉熵以避免进行 One Hot Encoding。

def soft1(x):
    return keras.activations.softmax(x, axis = -1)

def conv2d_block(input_tensor, n_filters, kernel_size = 3, batchnorm = True):

    x = Conv2D(filters = n_filters, kernel_size = (kernel_size, kernel_size),\
              kernel_initializer = 'he_normal', padding = 'same')(input_tensor)
    if batchnorm:
        x = BatchNormalization()(x)
    x = Activation('relu')(x)


    x = Conv2D(filters = n_filters, kernel_size = (kernel_size, kernel_size),\
              kernel_initializer = 'he_normal', padding = 'same')(input_tensor)
    if batchnorm:
        x = BatchNormalization()(x)
    x = Activation('relu')(x)

    return x

def …