keras.layers.TimeDistributed与hub.KerasLayer NotImplementedError [英] keras.layers.TimeDistributed with hub.KerasLayer NotImplementedError

问题描述

我想将tf.keras.TimeDistributed()层与最新TensorFLow V2版本(tf-nightly-gpu-2.0-preview)中的tf.hub inception_v3 CNN模型一起使用.输出如下所示.似乎tf.keras.TimeDistributed()并未完全实现与tf.hub模型一起使用.不知何故，无法计算输入层的形状.我的问题:是否有解决此问题的方法?

I want to use tf.keras.TimeDistributed() layer with the tf.hub inception_v3 CNN model from the latest TensorFLow V2 version (tf-nightly-gpu-2.0-preview). The output is shown below. It seemst that tf.keras.TimeDistributed() is not fully implemented to work with tf.hub models. Somehow, the shape of the input layer cannot be computed. My question: Is there a workaround this problem?

tf.keras.TimeDistributed可以正常工作.我只想将CNN模型应用于每个时间步骤.

tf.keras.TimeDistributed with regular tf.keras.layer works fine. I just would like to apply the CNN model to each time step.

import tensorflow as tf
import tensorflow_hub as hub 
from tensorflow.keras import layers, Model

model_url = "https://tfhub.dev/google/tf2- 

preview/inception_v3/feature_vector/3"

preview/inception_v3/feature_vector/3"

feature_layer = hub.KerasLayer(model_url,
                               input_shape = (299, 299, 3),
                               output_shape = [2048],
                               trainable = False)

video = layers.Input(shape = (None, 299, 299, 3))

encoded_frames = layers.TimeDistributed(feature_layer)(video)

model = Model(inputs = video, outputs = encoded_frames)

预期产量

tf.keras模型

Expected output

tf.keras model

文件"/usr/local/lib/python3.5/dist-packages/tensorflow/python/keras/engine/base_layer.py"，第489行，在compute_output_shape中 引发NotImplementedError NotImplementedError

File "/usr/local/lib/python3.5/dist-packages/tensorflow/python/keras/engine/base_layer.py", line 489, in compute_output_shape raise NotImplementedError NotImplementedError

推荐答案

在Tensorflow 2中，可以将自定义层与TimeDistributed层结合使用.引发错误是因为它无法计算输出形状(请参见

In Tensorflow 2 it is possible to use custom layers in combination with the TimeDistributed layer. The error is thrown because it can't compute the output shape (see here).

因此，在您的情况下，您应该能够继承KerasLayer的子类并手动实现compute_output_shape.

So in your case you should be able to subclass KerasLayer and implement compute_output_shape manually.

这篇关于keras.layers.TimeDistributed与hub.KerasLayer NotImplementedError的文章就介绍到这了，希望我们推荐的答案对大家有所帮助，也希望大家多多支持IT屋！