问题描述
我有以下 siamese 网络的网络定义:
def build_siamese_model(inputShape,embeddingDim=48):
# specify the inputs for the feature extractor network
inputs = Input(inputShape)
## first set of CONV => RELU => RESID=> POOL => DROPOUT layers
first_conv1 = Conv2D(32,(3,3),padding="same")(inputs)
first_batch_norm1=Batchnormalization()(first_conv1)
first_act1= LeakyReLU()(first_batch_norm1)
second_conv1 = Conv2D(32,(5,5),padding="same")(inputs)
second_batch_norm1=Batchnormalization()(second_conv1)
second_act1= LeakyReLU()(second_batch_norm1)
third_conv1 = Conv2D(32,(7,7),padding="same")(inputs)
third_batch_norm1=Batchnormalization()(third_conv1)
third_act1= LeakyReLU()(third_batch_norm1)
residual_block1= Add()([first_act1,second_act1,third_act1])
pool1 = MaxPooling2D(pool_size=(2,2))(residual_block1)
dropout1 = Dropout(0.3)(pool1)
#receiver Convolutional layer
receiver1_conv = Conv2D(32,padding="same")(dropout1)
receiver1_batch_norm=Batchnormalization()(receiver1_conv)
act_receiver1=LeakyReLU()(receiver1_batch_norm)
## second set of CONV => BN=> RELU => RESID=> POOL => DROPOUT layers
first_conv2 = Conv2D(32,padding="same")(act_receiver1)
first_batch_norm2=Batchnormalization()(first_conv2)
first_act2= LeakyReLU()(first_batch_norm2)
second_conv2 = Conv2D(32,padding="same")(act_receiver1)
second_batch_norm2=Batchnormalization()(second_conv2)
second_act2= LeakyReLU()(second_batch_norm2)
third_conv2 = Conv2D(32,padding="same")(act_receiver1)
third_batch_norm2=Batchnormalization()(third_conv2)
third_act2= LeakyReLU()(third_batch_norm2)
residual_block2= Add()([first_act2,second_act2,third_act2])
pool2 = MaxPooling2D(pool_size=(2,2))(residual_block2)
dropout2 = Dropout(0.3)(pool2)
#receiver Convolutional layer
receiver2_conv = Conv2D(32,padding="same")(dropout2)
receiver2_batch_norm=Batchnormalization()(receiver2_conv)
act_receiver2=LeakyReLU()(receiver2_batch_norm)
## last set of CONV => BN=> RELU => RESID=> POOL => DROPOUT layers
first_conv3 = Conv2D(32,padding="same")(act_receiver2)
first_batch_norm3=Batchnormalization()(first_conv3)
first_act3= LeakyReLU()(first_batch_norm3)
second_conv3 = Conv2D(32,padding="same")(act_receiver2)
second_batch_norm3=Batchnormalization()(second_conv3)
second_act3= LeakyReLU()(second_batch_norm3)
third_conv3 = Conv2D(32,padding="same")(act_receiver2)
third_batch_norm3=Batchnormalization()(third_conv3)
third_act3= LeakyReLU()(third_batch_norm3)
residual_block3= Add()([first_act3,second_act3,third_act3])
pool3 = MaxPooling2D(pool_size=(2,2))(residual_block3)
dropout3 = Dropout(0.3)(pool3)
#last receiver Convolutional layer
receiver3_conv = Conv2D(32,padding="same")(dropout3)
receiver3_batch_norm=Batchnormalization()(receiver3_conv)
act_receiver3=LeakyReLU()(receiver3_batch_norm)
# prepare the final outputs
pooledOutput = GlobalAveragePooling2D()(act_receiver3)
outputs = Dense(embeddingDim)(pooledOutput)
# build the model
model = Model(inputs,outputs)
return(model)
但是,这部分作为功能性 API 连接到我的网络的输入和输出。以下是我如何链接这些部分:
print("[INFO] building siamese network...")
imgA = Input(shape=config.IMG_SHAPE)
imgB = Input(shape=config.IMG_SHAPE)
featureExtractor = build_siamese_model(config.IMG_SHAPE)
featsA = featureExtractor(imgA)
featsB = featureExtractor(imgB)
distance = Lambda(utils.euclidean_distance)([featsA,featsB])
outputs = Dense(1,activation="sigmoid")(distance)
model = Model(inputs=[imgA,imgB],outputs=outputs)
但是,在编译模型时,这里是模型的摘要:
因此,我上面所做的网络定义似乎只是网络的一层。
那么,我想要什么?
我想加载模型并提取特定层的输出。特别是,我想要功能对象最后一层的输出(上面网络定义中的outputs = Dense(48)(pooledOutput))。这将为我在模型中测试的每对图像提供 48 个特征向量。
我尝试检查一些 Previous posts 并执行以下操作:
print("Step 1: Loading Model")
model1=load_model("where/the/model/is/located",compile=False)
#I tried the output of the firstlayer,for example
model_with_intermediate_layers = Model(inputs=model1.input,outputs = model1.layers[0].output)
pred = model_with_intermediate_layers.predict([pair_1,pair_2],steps = 1)
print(pred)
有什么问题?
上面代码的问题在于它只能访问来自 0、1、3 和 4 的层。0 和 1 给出输入形状,第 3 层给我分数,第 4 层是空的。 **我想访问中间层,尤其是特征提取器网络的最后一层。 ** 我该怎么做?
解决方法
考虑到 (i) 我的 Functional 对象是网络的第二层; (ii) 我想要它的最终层输出; (iii) 第二层的输出是第三层的输入,我用下面的代码解决了这个问题:
#I am getting layer's 3 input,which is the same as the second layer's output (last layer of my functional model)
model_intermediate = Model(inputs=model1.input,outputs = model1.layers[3].input)
#Here I get 2 48-d vectors.
pred_intermediate = model_intermediate.predict([pair_1,pair_2],steps = 1) # predict_generator is deprecated
pred_intermediate=np.array(pred_intermediate)
print(type(pred_intermediate))
print(pred_intermediate)
print(pred_intermediate.shape)
input()
它给了我我想要的