Styjun

Why didn't Nick Fury expose the villain's identity and plans?

Where should I connect my modem in this ethernet junction box?

How can I effectively communicate to recruiters that a phone call is not possible?

Find The One Element In An Array That is Different From The Others

Would dual wielding daggers be a viable choice for a covert bodyguard?

Print the last, middle and first character of your code

Why can a destructor change the state of a constant object?

Civil War story: man hanged from a bridge

Why are Hobbits so fond of mushrooms?

Machine learning and operations research projects

Do you know your 'KVZ's?

How to loop for 3 times in bash script when docker push fails?

How is angular momentum conserved for the orbiting body if the centripetal force disappears?

Can the Mage Hand cantrip be used to trip an enemy who is running away?

Has anyone in space seen or photographed a simple laser pointer from Earth?

Fivenum and a little bit

How to md5 a list of filepaths contained in a file?

definition of "percentile"

When did Tsar Nicholas II become Tsar of Russia?

Misspelling my name on my mathematical publications

Is a 10th-level Transmutation wizard considered a shapechanger for the purpose of effects such as Moonbeam?

Optimization terminology: "Exact" v. "Approximate"

How to find the object of reference of a latin relative pronoun?

What's the minimum number of sensors for a hobby GPS waypoint-following UAV?

Getting the filters values from CNN layers

How to get the current time in PythonHow do I sort a dictionary by value?Cannot make this autoencoder network function properly (with convolutional and maxpool layers)Convolutional Autoencoders: Black Feature MapsIssue with tf.nn.max_poolOutput of conv2d in kerasHow to get stride value in Conv2D layer Tensorflow?Collapsing consecutive linear layersCNN padding and stridingtensorflow: filters vs kernels and strides

.everyoneloves__top-leaderboard:empty,.everyoneloves__mid-leaderboard:empty,.everyoneloves__bot-mid-leaderboard:empty margin-bottom:0;

1

I have the following model (for example)

input_img = Input(shape=(224,224,1)) # size of the input image
x = Conv2D(64, (3, 3), strides=(1, 1), activation='relu', padding='same')(input_img)

I have several layers of such in my autoencoder model. I am particularly interested in the filters of the first layer. There are 64 filters each of size 3x3.

To get the filters, I tried using the following code:

x.layers[0].get_weights()[0]

but I am getting the error as follows:

AttributeError Traceback (most recent call last)
<ipython-input-166-96506292d6d7> in <module>()
 4 x = Conv2D(64, (3, 3), strides=(1, 1), activation='relu', padding='same')(input_img)
 5 
----> 6 x.layers[0].get_weights()[0]

AttributeError: 'Tensor' object has no attribute 'layers'

I am not using the sequential model. My model will be formed using the following command after several such layers.

 model = Model()

I am new to CNN and I don't even know if the get_weights function can help me get filters value. How do I get value of filters?

python tensorflow conv-neural-network keras-layer autoencoder

share|improve this question

edited Mar 26 at 2:30

enjal

asked Mar 26 at 2:18

enjalenjal

38511 gold badge88 silver badges2121 bronze badges

add a comment | 

1

I have the following model (for example)

input_img = Input(shape=(224,224,1)) # size of the input image
x = Conv2D(64, (3, 3), strides=(1, 1), activation='relu', padding='same')(input_img)

I have several layers of such in my autoencoder model. I am particularly interested in the filters of the first layer. There are 64 filters each of size 3x3.

To get the filters, I tried using the following code:

x.layers[0].get_weights()[0]

but I am getting the error as follows:

AttributeError Traceback (most recent call last)
<ipython-input-166-96506292d6d7> in <module>()
 4 x = Conv2D(64, (3, 3), strides=(1, 1), activation='relu', padding='same')(input_img)
 5 
----> 6 x.layers[0].get_weights()[0]

AttributeError: 'Tensor' object has no attribute 'layers'

I am not using the sequential model. My model will be formed using the following command after several such layers.

 model = Model()

I am new to CNN and I don't even know if the get_weights function can help me get filters value. How do I get value of filters?

python tensorflow conv-neural-network keras-layer autoencoder

share|improve this question

edited Mar 26 at 2:30

enjal

asked Mar 26 at 2:18

enjalenjal

38511 gold badge88 silver badges2121 bronze badges

add a comment | 

I have the following model (for example)

input_img = Input(shape=(224,224,1)) # size of the input image
x = Conv2D(64, (3, 3), strides=(1, 1), activation='relu', padding='same')(input_img)

I have several layers of such in my autoencoder model. I am particularly interested in the filters of the first layer. There are 64 filters each of size 3x3.

To get the filters, I tried using the following code:

x.layers[0].get_weights()[0]

but I am getting the error as follows:

AttributeError Traceback (most recent call last)
<ipython-input-166-96506292d6d7> in <module>()
 4 x = Conv2D(64, (3, 3), strides=(1, 1), activation='relu', padding='same')(input_img)
 5 
----> 6 x.layers[0].get_weights()[0]

AttributeError: 'Tensor' object has no attribute 'layers'

I am not using the sequential model. My model will be formed using the following command after several such layers.

 model = Model()

I am new to CNN and I don't even know if the get_weights function can help me get filters value. How do I get value of filters?

python tensorflow conv-neural-network keras-layer autoencoder

share|improve this question

edited Mar 26 at 2:30

enjal

asked Mar 26 at 2:18

enjalenjal

38511 gold badge88 silver badges2121 bronze badges

input_img = Input(shape=(224,224,1)) # size of the input image
x = Conv2D(64, (3, 3), strides=(1, 1), activation='relu', padding='same')(input_img)

x.layers[0].get_weights()[0]

AttributeError Traceback (most recent call last)
<ipython-input-166-96506292d6d7> in <module>()
 4 x = Conv2D(64, (3, 3), strides=(1, 1), activation='relu', padding='same')(input_img)
 5 
----> 6 x.layers[0].get_weights()[0]

AttributeError: 'Tensor' object has no attribute 'layers'

 model = Model()

share|improve this question

edited Mar 26 at 2:30

enjal

asked Mar 26 at 2:18

enjalenjal

38511 gold badge88 silver badges2121 bronze badges

share|improve this question

edited Mar 26 at 2:30

enjal

asked Mar 26 at 2:18

enjalenjal

38511 gold badge88 silver badges2121 bronze badges

asked Mar 26 at 2:18

enjalenjal

38511 gold badge88 silver badges2121 bronze badges

asked Mar 26 at 2:18

enjalenjal

38511 gold badge88 silver badges2121 bronze badges

1 Answer
1

active

oldest

votes

0

At the moment your code is calling the layers function on a layer definition itself.

The model first needs to be compiled and then you can use the layers function on the model to retrieve the weights of the specific layer.

In your case:

weights = model.layers[1].get_weights()

will give you the set of weights of the 1st convolutional layer

Which you can use after compiling the model:

model = Model(inputs=input_img, output=b)

Where b refers to the last layer in your model.

share|improve this answer

edited Mar 26 at 4:26

answered Mar 26 at 2:39

JimmyOnThePageJimmyOnThePage

54833 silver badges1515 bronze badges

• 
  
  
  

  
  

  
  
  
  
  
  

  
  
  The convolutional layer can be accessed through the index [1] in the call to get_weights, since the first element will now be your input layer
  
  – JimmyOnThePage
  Mar 26 at 2:45
  

  
  
  
  


• 
  
  
  

  
  

  
  
  
  
  
  

  
  
  It gives me an error IndexError: list index out of range. Also, I want to get the filters on the first layer to apply geometric translations on those filters. I don't think this is possible after I compile it. What do you think about it?
  
  – enjal
  Mar 26 at 4:11
  
  

  
  
  
  


• 
  
  
  

  
  

  
  
  
  
  
  

  
  
  As mentioned in the comment, calling the function on the input layer will give you that error, since the input layer does not have weights.
  
  – JimmyOnThePage
  Mar 26 at 4:18
  

  
  
  
  


• 
  
  
  

  
  

  
  
  
  
  
  

  
  
  Not sure I understand the second part of your comment. Are you attempting to create a layer in the model that is a 'geometric translation' (you will have to clarify on this) of the first convolutional layer? You do realise the model is initially compiled with random weights, and first needs to be trained? Because I see no reason to apply transformations to filters containing random weights
  
  – JimmyOnThePage
  Mar 26 at 4:21
  

  
  
  
  


• 
  
  
  

  
  

  
  
  
  
  
  

  
  
  I did it after I compiled my model the same way you showed in the answer but still got the error.
  
  – enjal
  Mar 26 at 4:47
  

  
  
  
  

 | 
show 5 more comments

Your Answer

StackExchange.ifUsing("editor", function ()
StackExchange.using("externalEditor", function ()
StackExchange.using("snippets", function ()
StackExchange.snippets.init();
);
);
, "code-snippets");

StackExchange.ready(function()
var channelOptions =
tags: "".split(" "),
id: "1"
;
initTagRenderer("".split(" "), "".split(" "), channelOptions);

StackExchange.using("externalEditor", function()
// Have to fire editor after snippets, if snippets enabled
if (StackExchange.settings.snippets.snippetsEnabled)
StackExchange.using("snippets", function()
createEditor();
);

else
createEditor();

);

function createEditor()
StackExchange.prepareEditor(
heartbeatType: 'answer',
autoActivateHeartbeat: false,
convertImagesToLinks: true,
noModals: true,
showLowRepImageUploadWarning: true,
reputationToPostImages: 10,
bindNavPrevention: true,
postfix: "",
imageUploader:
brandingHtml: "Powered by u003ca class="icon-imgur-white" href="https://imgur.com/"u003eu003c/au003e",
contentPolicyHtml: "User contributions licensed under u003ca href="https://creativecommons.org/licenses/by-sa/3.0/"u003ecc by-sa 3.0 with attribution requiredu003c/au003e u003ca href="https://stackoverflow.com/legal/content-policy"u003e(content policy)u003c/au003e",
allowUrls: true
,
onDemand: true,
discardSelector: ".discard-answer"
,immediatelyShowMarkdownHelp:true
);



);

draft saved

draft discarded

Sign up or log in

StackExchange.ready(function ()
StackExchange.helpers.onClickDraftSave('#login-link');
);

Sign up using Google

Sign up using Facebook

Sign up using Email and Password

Post as a guest

Name

Email

Required, but never shown

StackExchange.ready(
function ()
StackExchange.openid.initPostLogin('.new-post-login', 'https%3a%2f%2fstackoverflow.com%2fquestions%2f55348940%2fgetting-the-filters-values-from-cnn-layers%23new-answer', 'question_page');

);

Post as a guest

Name

Email

Required, but never shown

0

At the moment your code is calling the layers function on a layer definition itself.

The model first needs to be compiled and then you can use the layers function on the model to retrieve the weights of the specific layer.

In your case:

weights = model.layers[1].get_weights()

will give you the set of weights of the 1st convolutional layer

Which you can use after compiling the model:

model = Model(inputs=input_img, output=b)

Where b refers to the last layer in your model.

share|improve this answer

edited Mar 26 at 4:26

answered Mar 26 at 2:39

JimmyOnThePageJimmyOnThePage

54833 silver badges1515 bronze badges

• 
  
  
  

  
  

  
  
  
  
  
  

  
  
  The convolutional layer can be accessed through the index [1] in the call to get_weights, since the first element will now be your input layer
  
  – JimmyOnThePage
  Mar 26 at 2:45
  

  
  
  
  


• 
  
  
  

  
  

  
  
  
  
  
  

  
  
  It gives me an error IndexError: list index out of range. Also, I want to get the filters on the first layer to apply geometric translations on those filters. I don't think this is possible after I compile it. What do you think about it?
  
  – enjal
  Mar 26 at 4:11
  
  

  
  
  
  


• 
  
  
  

  
  

  
  
  
  
  
  

  
  
  As mentioned in the comment, calling the function on the input layer will give you that error, since the input layer does not have weights.
  
  – JimmyOnThePage
  Mar 26 at 4:18
  

  
  
  
  


• 
  
  
  

  
  

  
  
  
  
  
  

  
  
  Not sure I understand the second part of your comment. Are you attempting to create a layer in the model that is a 'geometric translation' (you will have to clarify on this) of the first convolutional layer? You do realise the model is initially compiled with random weights, and first needs to be trained? Because I see no reason to apply transformations to filters containing random weights
  
  – JimmyOnThePage
  Mar 26 at 4:21
  

  
  
  
  


• 
  
  
  

  
  

  
  
  
  
  
  

  
  
  I did it after I compiled my model the same way you showed in the answer but still got the error.
  
  – enjal
  Mar 26 at 4:47
  

  
  
  
  

 | 
show 5 more comments

0

At the moment your code is calling the layers function on a layer definition itself.

The model first needs to be compiled and then you can use the layers function on the model to retrieve the weights of the specific layer.

In your case:

weights = model.layers[1].get_weights()

will give you the set of weights of the 1st convolutional layer

Which you can use after compiling the model:

model = Model(inputs=input_img, output=b)

Where b refers to the last layer in your model.

share|improve this answer

edited Mar 26 at 4:26

answered Mar 26 at 2:39

JimmyOnThePageJimmyOnThePage

54833 silver badges1515 bronze badges

• 
  
  
  

  
  

  
  
  
  
  
  

  
  
  The convolutional layer can be accessed through the index [1] in the call to get_weights, since the first element will now be your input layer
  
  – JimmyOnThePage
  Mar 26 at 2:45
  

  
  
  
  


• 
  
  
  

  
  

  
  
  
  
  
  

  
  
  It gives me an error IndexError: list index out of range. Also, I want to get the filters on the first layer to apply geometric translations on those filters. I don't think this is possible after I compile it. What do you think about it?
  
  – enjal
  Mar 26 at 4:11
  
  

  
  
  
  


• 
  
  
  

  
  

  
  
  
  
  
  

  
  
  As mentioned in the comment, calling the function on the input layer will give you that error, since the input layer does not have weights.
  
  – JimmyOnThePage
  Mar 26 at 4:18
  

  
  
  
  


• 
  
  
  

  
  

  
  
  
  
  
  

  
  
  Not sure I understand the second part of your comment. Are you attempting to create a layer in the model that is a 'geometric translation' (you will have to clarify on this) of the first convolutional layer? You do realise the model is initially compiled with random weights, and first needs to be trained? Because I see no reason to apply transformations to filters containing random weights
  
  – JimmyOnThePage
  Mar 26 at 4:21
  

  
  
  
  


• 
  
  
  

  
  

  
  
  
  
  
  

  
  
  I did it after I compiled my model the same way you showed in the answer but still got the error.
  
  – enjal
  Mar 26 at 4:47
  

  
  
  
  

 | 
show 5 more comments

At the moment your code is calling the layers function on a layer definition itself.

The model first needs to be compiled and then you can use the layers function on the model to retrieve the weights of the specific layer.

In your case:

weights = model.layers[1].get_weights()

will give you the set of weights of the 1st convolutional layer

Which you can use after compiling the model:

model = Model(inputs=input_img, output=b)

Where b refers to the last layer in your model.

share|improve this answer

edited Mar 26 at 4:26

answered Mar 26 at 2:39

JimmyOnThePageJimmyOnThePage

54833 silver badges1515 bronze badges

weights = model.layers[1].get_weights()

model = Model(inputs=input_img, output=b)

share|improve this answer

edited Mar 26 at 4:26

answered Mar 26 at 2:39

JimmyOnThePageJimmyOnThePage

54833 silver badges1515 bronze badges

answered Mar 26 at 2:39

JimmyOnThePageJimmyOnThePage

54833 silver badges1515 bronze badges

answered Mar 26 at 2:39

JimmyOnThePageJimmyOnThePage

54833 silver badges1515 bronze badges