Keras pretrain CNN with TimeDistributed

时间:2017-02-18 09:28:57

标签: python tensorflow keras recurrent-neural-network resnet

这是我的问题,我想在TimeDistributed层中使用pretrain CNN网络之一。但我实施它有一些问题。

这是我的模特:

def bnn_model(max_len):
    # sequence length and resnet input size
    x = Input(shape=(maxlen, 224, 224, 3))

    base_model = ResNet50.ResNet50(weights='imagenet',  include_top=False)

    for layer in base_model.layers:
        layer.trainable = False

    som = TimeDistributed(base_model)(x)

    #the ouput of the model is [1, 1, 2048], need to squeeze
    som = Lambda(lambda x: K.squeeze(K.squeeze(x,2),2))(som)

    bnn = Bidirectional(LSTM(300))(som)
    bnn = Dropout(0.5)(bnn)

    pred = Dense(1, activation='sigmoid')(bnn)

    model = Model(input=x, output=pred)

    model.compile(optimizer=Adam(lr=1.0e-5), loss="mse", metrics=["accuracy"])

    return model

编译模型时,我没有错误。但是当我开始训练时,我得到以下错误:

tensorflow/core/framework/op_kernel.cc:975] Invalid argument: You must feed a value for placeholder tensor 'input_2' with dtype float
[[Node: input_2 = Placeholder[dtype=DT_FLOAT, shape=[], _device="/job:localhost/replica:0/task:0/gpu:0"]()]]

我检查过并且确实发送了float32但是对于input1,input2是pretrain Resnet中的输入。

这里只是概述模型摘要。 (注意:它并不奇怪,它没有显示Resnet内部发生的事情,但从不介意)

____________________________________________________________________________________________________
Layer (type)                     Output Shape          Param #     Connected to                     
====================================================================================================
input_1 (InputLayer)             (None, 179, 224, 224, 0                                            
____________________________________________________________________________________________________
timedistributed_1 (TimeDistribut (None, 179, 1, 1, 204 23587712    input_1[0][0]                    
____________________________________________________________________________________________________
lambda_1 (Lambda)                (None, 179, 2048)     0           timedistributed_1[0][0]          
____________________________________________________________________________________________________
bidirectional_1 (Bidirectional)  (None, 600)           5637600     lambda_1[0][0]                   
____________________________________________________________________________________________________
dropout_1 (Dropout)              (None, 600)           0           bidirectional_1[0][0]            
____________________________________________________________________________________________________
dense_1 (Dense)                  (None, 1)             601         dropout_1[0][0]                  
====================================================================================================
Total params: 29,225,913
Trainable params: 5,638,201
Non-trainable params: 23,587,712
____________________________________________________________________________________________________

我猜我没有正确使用TimeDistributed,我看到没有人试图这样做。我希望有人可以指导我。

编辑:

问题来自于ResNet50.ResNet50(weights='imagenet', include_top=False)在图表中创建自己的输入。

所以我想我需要做ResNet50.ResNet50(weights='imagenet', input_tensor=x, include_top=False)之类的事情,但我不知道如何将其与TimeDistributed结合起来。

我试过

base_model = Lambda(lambda x : ResNet50.ResNet50(weights='imagenet',  input_tensor=x, include_top=False))
som = TimeDistributed(base_model)(in_ten)

但它不起作用。

2 个答案:

答案 0 :(得分:2)

我的快速解决方案有点难看。

我刚刚复制了ResNet的代码并将TimeDistributed添加到了所有图层,然后从" basic"中加载了权重。 ResNet在我定制的ResNet上。

注意:

为了能够像这样分析图像序列,需要在gpu上占用大量内存。

答案 1 :(得分:1)

考虑到您使用的是来自keras的预训练网络,您也可以将其替换为自己的预训练网络。

这是一个简单的解决方案::

model_vgg=keras.applications.VGG16(input_shape=(256, 256, 3),
                                           include_top=False,
                                           weights='imagenet')
model_vgg.trainable = False
model_vgg.summary()

如果您想使用任何中间层,则将“ block2_pool”替换为最后一层的名称::

intermediate_model= Model(inputs=model_vgg.input, outputs=model_vgg.get_layer('block2_pool').output)
intermediate_model.summary()

最后将其包装在TimeDistributed层中

input_tensor = Input(shape=(time_steps,height, width, channels))
timeDistributed_layer = TimeDistributed( intermediate_model )(input_tensor)

现在您可以轻松地做到::

my_time_model = Model( inputs = input_tensor, outputs = timeDistributed_layer )