Keras 模型的验证损失与损失函数输出不匹配

Question

我正在从事生物医学图像分割项目。我正在使用 U-net 模型来完成这项工作。问题是，当我训练模型时，验证损失似乎并不实用。

我使用 dice_coef_loss 作为损失函数，以及指标。训练结果如下图。该图清楚地表明验证损失不遵循我的损失函数，因为这两个图是可区分的。但是，train loss 确实遵循 train dice_coef_loss 值。

（左起第一个图像是训练和验证损失，第三个是tarining和验证dice_coef_loss作为度量）

The history graph of training

（抱歉，我还没有资格嵌入图片，请检查 link）

这是我的模型

def unet(input_size=(256,256,1)):
    inputs = Input(input_size)

    conv1 = Conv2D(32, (3, 3), activation='relu', padding='same')(inputs)
    conv1 = Conv2D(32, (3, 3), activation='relu', padding='same')(conv1)
    pool1 = MaxPooling2D(pool_size=(2, 2))(conv1)

    conv2 = Conv2D(64, (3, 3), activation='relu', padding='same')(pool1)
    conv2 = Conv2D(64, (3, 3), activation='relu', padding='same')(conv2)
    pool2 = MaxPooling2D(pool_size=(2, 2))(conv2)

    conv3 = Conv2D(128, (3, 3), activation='relu', padding='same')(pool2)
    conv3 = Conv2D(128, (3, 3), activation='relu', padding='same')(conv3)
    pool3 = MaxPooling2D(pool_size=(2, 2))(conv3)

    conv4 = Conv2D(256, (3, 3), activation='relu', padding='same')(pool3)
    conv4 = Conv2D(256, (3, 3), activation='relu', padding='same')(conv4)
    pool4 = MaxPooling2D(pool_size=(2, 2))(conv4)

    conv5 = Conv2D(512, (3, 3), activation='relu', padding='same')(pool4)
    conv5 = Conv2D(512, (3, 3), activation='relu', padding='same')(conv5)

    up6 = concatenate([Conv2DTranspose(256, (2, 2), strides=(2, 2), padding='same')(conv5), conv4], axis=3)
    conv6 = Conv2D(256, (3, 3), activation='relu', padding='same')(up6)
    conv6 = Conv2D(256, (3, 3), activation='relu', padding='same')(conv6)

    up7 = concatenate([Conv2DTranspose(128, (2, 2), strides=(2, 2), padding='same')(conv6), conv3], axis=3)
    conv7 = Conv2D(128, (3, 3), activation='relu', padding='same')(up7)
    conv7 = Conv2D(128, (3, 3), activation='relu', padding='same')(conv7)

    up8 = concatenate([Conv2DTranspose(64, (2, 2), strides=(2, 2), padding='same')(conv7), conv2], axis=3)
    conv8 = Conv2D(64, (3, 3), activation='relu', padding='same')(up8)
    conv8 = Conv2D(64, (3, 3), activation='relu', padding='same')(conv8)

    up9 = concatenate([Conv2DTranspose(32, (2, 2), strides=(2, 2), padding='same')(conv8), conv1], axis=3)
    conv9 = Conv2D(32, (3, 3), activation='relu', padding='same')(up9)
    conv9 = Conv2D(32, (3, 3), activation='relu', padding='same')(conv9)

    conv10 = Conv2D(1, (1, 1), activation='sigmoid')(conv9)

    return Model(inputs=[inputs], outputs=[conv10])

这是损失函数

import keras.backend as K
smooth=100

def dice_coef(y_true, y_pred):
    y_truef=K.flatten(y_true)
    y_predf=K.flatten(y_pred)
    And=K.sum(y_truef* y_predf)
    return((2* And + smooth) / (K.sum(y_truef) + K.sum(y_predf) + smooth))

def dice_coef_loss(y_true, y_pred):
    return -dice_coef(y_true, y_pred)

编译

model.compile(optimizer=Adam(lr=1e-5), loss=dice_coef_loss, metrics=["binary_accuracy", dice_coef_loss, iou])

备注

我试过：

学习率调整
调整批量大小
数据扩充
改变损失函数

如果有人想看这里的代码是link to kaggle kernel

补充说明

为了让我自己清楚:

第一个和第二个图是使用相同数据集（来自训练或验证时间）的相同函数生成的，所以我希望第一个图的 train_loss 曲线等于 train_dice_coef_loss 来自第三个情节。第一个图的 val_loss 曲线等于第三个图的 val_dice_coef_loss。

但不幸的是 val_loss 曲线与 val_dice_coef_loss 曲线不匹配。

P.S。我刚来这地方。感谢任何改进我的问题的建议。谢谢。

Answer 1

因为，ZabirAlNazi 表明这是所用库的问题。将导入从 keras 更改为 tensorflow.keras 解决了该问题。

from tensorflow.keras.models import Model, load_model, save_model
from tensorflow.keras.layers import Input, Activation, BatchNormalization, Dropout, Lambda, Dense, Conv2D

Keras 模型的验证损失与损失函数输出不匹配

Keras model's validation loss doesn't match the loss function output

python

image-segmentation

deep-learning

keras

这是我的模型

这是损失函数

编译

备注

补充说明