使用 keras sequential API 提取嵌套层特征（从预训练模型）

Question

我有以下用于迁移学习的简单模型，使用预训练模型 (VGG16)，没有 FC 层，然后是一些新层，定义为 keras 顺序 API。

IMG_SHAPE = (224, 224, 3)

# vgg16
pretrained_model = tf.keras.applications.vgg16.VGG16(
        weights='imagenet', 
        include_top=False, 
        input_shape=IMG_SHAPE,
    )
# freeze pretrained layers
pretrained_model.trainable = False

model = tf.keras.Sequential([
    pretrained_model,
    tf.keras.layers.BatchNormalization(),
    tf.keras.layers.Flatten(),
    tf.keras.layers.Dense(3, activation='softmax'),
])

请注意，模型摘要不显示 VGG16 的内部层：

model.summary()

#Model: "sequential"
#_________________________________________________________________
# Layer (type)                Output Shape              Param #   
#=================================================================
# vgg16 (Functional)          (None, 4, 4, 512)         14714688  
#                                                                 
# batch_normalization (BatchN  (None, 4, 4, 512)        2048      
# ormalization)                                                   
#                                                                 
# flatten (Flatten)           (None, 8192)              0         
                                                                 
# dense (Dense)               (None, 2)                 16386     
                                                                 
#=================================================================
#Total params: 14,733,122
#Trainable params: 17,410
#Non-trainable params: 14,715,712

我已经在我的自定义数据集上训练了上述模型，并通过迁移学习在我的测试数据集上获得了所需的准确性。

现在，假设我想创建一个新模型（例如，计算激活图）接受输入作为前一个模型的输入，作为输出，我想要一个中间输出（通过提取预训练模型的卷积层的特征，例如 block5_conv3）以及前一个模型的输出.那是我陷入困境并且遇到错误的地方。例如，我定义了如下新模型：

grad_model = tf.keras.models.Model(
        [pretrained_model.inputs], 
        [pretrained_model.get_layer('block5_conv3').output, model.output]
    )

出现以下错误：

ValueError: Graph disconnected: cannot obtain value for tensor KerasTensor(type_spec=TensorSpec(shape=(None, 150, 150, 3), dtype=tf.float32, name='vgg16_input'), name='vgg16_input', description="created by layer 'vgg16_input'") at layer "vgg16". The following previous layers were accessed without issue: ['block1_conv1', 'block1_conv2', 'block1_pool', 'block2_conv1', 'block2_conv2', 'block2_pool', 'block3_conv1', 'block3_conv2', 'block3_conv3', 'block3_pool', 'block4_conv1', 'block4_conv2', 'block4_conv3']

或点赞：

grad_model = tf.keras.models.Model(
        [model.inputs], 
        [pretrained_model.get_layer('block5_conv3').output, model.output]
    )

出现以下错误：

ValueError: Graph disconnected: cannot obtain value for tensor KerasTensor(type_spec=TensorSpec(shape=(None, 150, 150, 3), dtype=tf.float32, name='input_1'), name='input_1', description="created by layer 'vgg16'") at layer "block1_conv1". The following previous layers were accessed without issue: []

我也试过设置模型的输入层名称和嵌套在里面的预训练模型的名称，使输入层名称相同：

pretrained_model.layers[0]._name = model.layers[0]._name

但出现同样的错误。

我认为可以更改模型结构（例如，使用 keras 函数式 API 等）来定义 grad_model，但不确定如何更改。另外，我更想知道是否有一种方法可以在不更改模型结构/不需要我重新训练的情况下解决问题。

Answer 1

到目前为止，根据@M.Innat的评论，我可以使用keras功能API（注意参数数量保持不变）和re-training解决问题:

inputs = tf.keras.Input(shape=IMG_SHAPE)

x = keras.applications.xception.Xception(
    input_tensor=inputs, 
    include_top=False,
    weights='imagenet'
)

x.trainable = False

x = tf.keras.layers.BatchNormalization()(x.output) 
x = tf.keras.layers.Flatten()(x)
x = tf.keras.layers.Dense(2, activation='softmax')(x)

model = tf.keras.Model(inputs, x)

model.summary()

# Model: "model_1262"
#_________________________________________________________________
# Layer (type)                Output Shape              Param #   
#=================================================================
# input_3 (InputLayer)        [(None, 150, 150, 3)]     0        
#                                                                 
# block1_conv1 (Conv2D)       (None, 150, 150, 64)      1792      
#                                                                 
# block1_conv2 (Conv2D)       (None, 150, 150, 64)      36928     
#                                                                 
# block1_pool (MaxPooling2D)  (None, 75, 75, 64)        0         
#                                                                 
# block2_conv1 (Conv2D)       (None, 75, 75, 128)       73856     
#                                                                 
# block2_conv2 (Conv2D)       (None, 75, 75, 128)       147584    
#                                                                 
# block2_pool (MaxPooling2D)  (None, 37, 37, 128)       0         
#                                                                 
# block3_conv1 (Conv2D)       (None, 37, 37, 256)       295168    
#                                                                 
# block3_conv2 (Conv2D)       (None, 37, 37, 256)       590080    
#                                                                 
# block3_conv3 (Conv2D)       (None, 37, 37, 256)       590080    
#                                                                 
# block3_pool (MaxPooling2D)  (None, 18, 18, 256)       0         
#                                                                 
# block4_conv1 (Conv2D)       (None, 18, 18, 512)       1180160   
#                                                                 
# block4_conv2 (Conv2D)       (None, 18, 18, 512)       2359808   
#                                                                 
# block4_conv3 (Conv2D)       (None, 18, 18, 512)       2359808   
#                                                                 
# block4_pool (MaxPooling2D)  (None, 9, 9, 512)         0         
#                                                                 
# block5_conv1 (Conv2D)       (None, 9, 9, 512)         2359808   
#                                                                 
# block5_conv2 (Conv2D)       (None, 9, 9, 512)         2359808   
#                                                                 
# block5_conv3 (Conv2D)       (None, 9, 9, 512)         2359808   
#                                                                 
# block5_pool (MaxPooling2D)  (None, 4, 4, 512)         0         
#                                                                 
# batch_normalization_9 (Batc  (None, 4, 4, 512)        2048      
# hNormalization)                                                 
#                                                                 
# flatten_1 (Flatten)         (None, 8192)              0         
#                                                                 
# dense_1 (Dense)             (None, 2)                 16386     
#                                                                 
#=================================================================
#Total params: 14,733,122
#Trainable params: 17,410
#Non-trainable params: 14,715,712
#_________________________________________________________________

以及以下用于提取 class 激活图的中间特征的代码有效：

grad_model = tf.keras.models.Model(
        [model.inputs], [model.get_layer('block5_conv3').output, model.output] 
    )

使用 keras sequential API 提取嵌套层特征（从预训练模型）

Extracting nested layer features (from a pretrained model) with keras sequential API

python

computer-vision

deep-learning

keras

tensorflow