How to investigate what has been learnt in CNNs?
$begingroup$
My CNN was trained using close-up images of dogs. For testing, we input an image with a dog next to its owner and it was classified as a dog. How is that possible?
Isn't this image a different vector?
machine-learning neural-network deep-learning cnn image-classification
edited 4 hours ago
Media
6,76552057
asked 20 hours ago
πTh0nπTh0n
112
New contributor
πTh0n is a new contributor to this site. Take care in asking for clarification, commenting, and answering.
Check out our Code of Conduct.
$endgroup$
add a comment |
$begingroup$
My CNN was trained using close-up images of dogs. For testing, we input an image with a dog next to its owner and it was classified as a dog. How is that possible?
Isn't this image a different vector?
machine-learning neural-network deep-learning cnn image-classification
edited 4 hours ago
Media
6,76552057
asked 20 hours ago
πTh0nπTh0n
112
New contributor
πTh0n is a new contributor to this site. Take care in asking for clarification, commenting, and answering.
Check out our Code of Conduct.
$endgroup$
$begingroup$
Please add more information what are you trying to do. E.g. Are you doing image classification or semantic segmentation?
$endgroup$
– Antonio Jurić
11 hours ago
$begingroup$
I'm doing image classification using fastai module (cnn built upon resnet34 architecture)
$endgroup$
– πTh0n
7 hours ago
add a comment |
$begingroup$
My CNN was trained using close-up images of dogs. For testing, we input an image with a dog next to its owner and it was classified as a dog. How is that possible?
Isn't this image a different vector?
machine-learning neural-network deep-learning cnn image-classification
edited 4 hours ago
Media
6,76552057
asked 20 hours ago
πTh0nπTh0n
112
New contributor
πTh0n is a new contributor to this site. Take care in asking for clarification, commenting, and answering.
Check out our Code of Conduct.
$endgroup$
My CNN was trained using close-up images of dogs. For testing, we input an image with a dog next to its owner and it was classified as a dog. How is that possible?
Isn't this image a different vector?
machine-learning neural-network deep-learning cnn image-classification
machine-learning neural-network deep-learning cnn image-classification
edited 4 hours ago
Media
6,76552057
asked 20 hours ago
πTh0nπTh0n
112
New contributor
πTh0n is a new contributor to this site. Take care in asking for clarification, commenting, and answering.
Check out our Code of Conduct.
edited 4 hours ago
Media
6,76552057
asked 20 hours ago
πTh0nπTh0n
112
New contributor
πTh0n is a new contributor to this site. Take care in asking for clarification, commenting, and answering.
Check out our Code of Conduct.
edited 4 hours ago
Media
6,76552057
edited 4 hours ago
Media
6,76552057
edited 4 hours ago
Media
6,76552057
6,76552057
asked 20 hours ago
πTh0nπTh0n
112
New contributor
πTh0n is a new contributor to this site. Take care in asking for clarification, commenting, and answering.
Check out our Code of Conduct.
asked 20 hours ago
πTh0nπTh0n
112
asked 20 hours ago
πTh0nπTh0n
112
112
New contributor
πTh0n is a new contributor to this site. Take care in asking for clarification, commenting, and answering.
Check out our Code of Conduct.
New contributor
πTh0n is a new contributor to this site. Take care in asking for clarification, commenting, and answering.
Check out our Code of Conduct.
πTh0n is a new contributor to this site. Take care in asking for clarification, commenting, and answering.
Check out our Code of Conduct.
$begingroup$
Please add more information what are you trying to do. E.g. Are you doing image classification or semantic segmentation?
$endgroup$
– Antonio Jurić
11 hours ago
$begingroup$
I'm doing image classification using fastai module (cnn built upon resnet34 architecture)
$endgroup$
– πTh0n
7 hours ago
add a comment |
$begingroup$
Please add more information what are you trying to do. E.g. Are you doing image classification or semantic segmentation?
$endgroup$
– Antonio Jurić
11 hours ago
$begingroup$
I'm doing image classification using fastai module (cnn built upon resnet34 architecture)
$endgroup$
– πTh0n
7 hours ago
$begingroup$
Please add more information what are you trying to do. E.g. Are you doing image classification or semantic segmentation?
$endgroup$
– Antonio Jurić
11 hours ago
$begingroup$
Please add more information what are you trying to do. E.g. Are you doing image classification or semantic segmentation?
$endgroup$
– Antonio Jurić
11 hours ago
$begingroup$
I'm doing image classification using fastai module (cnn built upon resnet34 architecture)
$endgroup$
– πTh0n
7 hours ago
$begingroup$
I'm doing image classification using fastai module (cnn built upon resnet34 architecture)
$endgroup$
– πTh0n
7 hours ago
add a comment |
1 Answer
1
active
oldest
votes
$begingroup$
There is a famous paper about tracking what a CNN network has learnt. You can visualise to see which parts are more engaged in the classification using DeConvNet. In this paper, it was officially observed that first layers attempt to find simple lines and edges while deeper layers try to put the previous things together to make abstract concepts, like mouth, eye and such meaningful things. As an example take a look at the following image:
I guess there are implementations of this paper that you can't replace your pre-trained model with the one already exists and see what exactly is learnt by your network.
Isn't this image a different vector?
They are different vectors. ML and DL models are for generalisation which means they should be good at test time.
answered 4 hours ago
MediaMedia
6,76552057
$endgroup$
add a comment |
Your Answer
StackExchange.ifUsing("editor", function () {
return StackExchange.using("mathjaxEditing", function () {
StackExchange.MarkdownEditor.creationCallbacks.add(function (editor, postfix) {
StackExchange.mathjaxEditing.prepareWmdForMathJax(editor, postfix, [["$", "$"], ["\\(","\\)"]]);
});
});
}, "mathjax-editing");
StackExchange.ready(function() {
var channelOptions = {
tags: "".split(" "),
id: "557"
};
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: false,
noModals: true,
showLowRepImageUploadWarning: true,
reputationToPostImages: null,
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
});
}
});
πTh0n is a new contributor. Be nice, and check out our Code of Conduct.
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
Required, but never shown
StackExchange.ready(
function () {
StackExchange.openid.initPostLogin('.new-post-login', 'https%3a%2f%2fdatascience.stackexchange.com%2fquestions%2f44673%2fhow-to-investigate-what-has-been-learnt-in-cnns%23new-answer', 'question_page');
}
);
Post as a guest
Required, but never shown
1 Answer
1
active
oldest
votes
1 Answer
1
active
oldest
votes
active
oldest
votes
active
oldest
votes
$begingroup$
There is a famous paper about tracking what a CNN network has learnt. You can visualise to see which parts are more engaged in the classification using DeConvNet. In this paper, it was officially observed that first layers attempt to find simple lines and edges while deeper layers try to put the previous things together to make abstract concepts, like mouth, eye and such meaningful things. As an example take a look at the following image:
I guess there are implementations of this paper that you can't replace your pre-trained model with the one already exists and see what exactly is learnt by your network.
Isn't this image a different vector?
They are different vectors. ML and DL models are for generalisation which means they should be good at test time.
answered 4 hours ago
MediaMedia
6,76552057
$endgroup$
add a comment |
$begingroup$
There is a famous paper about tracking what a CNN network has learnt. You can visualise to see which parts are more engaged in the classification using DeConvNet. In this paper, it was officially observed that first layers attempt to find simple lines and edges while deeper layers try to put the previous things together to make abstract concepts, like mouth, eye and such meaningful things. As an example take a look at the following image:
I guess there are implementations of this paper that you can't replace your pre-trained model with the one already exists and see what exactly is learnt by your network.
Isn't this image a different vector?
They are different vectors. ML and DL models are for generalisation which means they should be good at test time.
answered 4 hours ago
MediaMedia
6,76552057
$endgroup$
add a comment |
$begingroup$
There is a famous paper about tracking what a CNN network has learnt. You can visualise to see which parts are more engaged in the classification using DeConvNet. In this paper, it was officially observed that first layers attempt to find simple lines and edges while deeper layers try to put the previous things together to make abstract concepts, like mouth, eye and such meaningful things. As an example take a look at the following image:
I guess there are implementations of this paper that you can't replace your pre-trained model with the one already exists and see what exactly is learnt by your network.
Isn't this image a different vector?
They are different vectors. ML and DL models are for generalisation which means they should be good at test time.
answered 4 hours ago
MediaMedia
6,76552057
$endgroup$
There is a famous paper about tracking what a CNN network has learnt. You can visualise to see which parts are more engaged in the classification using DeConvNet. In this paper, it was officially observed that first layers attempt to find simple lines and edges while deeper layers try to put the previous things together to make abstract concepts, like mouth, eye and such meaningful things. As an example take a look at the following image:
I guess there are implementations of this paper that you can't replace your pre-trained model with the one already exists and see what exactly is learnt by your network.
Isn't this image a different vector?
They are different vectors. ML and DL models are for generalisation which means they should be good at test time.
answered 4 hours ago
MediaMedia
6,76552057
edited 4 hours ago
answered 4 hours ago
MediaMedia
6,76552057
answered 4 hours ago
MediaMedia
6,76552057
answered 4 hours ago
MediaMedia
6,76552057
6,76552057
add a comment |
add a comment |
πTh0n is a new contributor. Be nice, and check out our Code of Conduct.
πTh0n is a new contributor. Be nice, and check out our Code of Conduct.
πTh0n is a new contributor. Be nice, and check out our Code of Conduct.
πTh0n is a new contributor. Be nice, and check out our Code of Conduct.
Thanks for contributing an answer to Data Science Stack Exchange!
- Please be sure to answer the question. Provide details and share your research!
But avoid …
- Asking for help, clarification, or responding to other answers.
- Making statements based on opinion; back them up with references or personal experience.
Use MathJax to format equations. MathJax reference.
To learn more, see our tips on writing great answers.
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
Required, but never shown
StackExchange.ready(
function () {
StackExchange.openid.initPostLogin('.new-post-login', 'https%3a%2f%2fdatascience.stackexchange.com%2fquestions%2f44673%2fhow-to-investigate-what-has-been-learnt-in-cnns%23new-answer', 'question_page');
}
);
Post as a guest
Required, but never shown
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
Required, but never shown
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
Required, but never shown
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
Sign up using Google
Sign up using Facebook
Sign up using Email and Password
Post as a guest
Required, but never shown
Required, but never shown
Required, but never shown
Required, but never shown
Required, but never shown
Required, but never shown
Required, but never shown
Required, but never shown
Required, but never shown
$begingroup$
Please add more information what are you trying to do. E.g. Are you doing image classification or semantic segmentation?
$endgroup$
– Antonio Jurić
11 hours ago
$begingroup$
I'm doing image classification using fastai module (cnn built upon resnet34 architecture)
$endgroup$
– πTh0n
7 hours ago