Talk:Convolutional neural network: Difference between revisions

Revision as of 16:51, 14 December 2020

Latest revision as of 00:10, 8 May 2025

(31 intermediate revisions by 21 users not shown)

Line 1:

{{talkheaderTalk header}}

{{WikiProject Computing|class=C|importance=low|needs-image=yes}}

{{annual readership|scale=log}}

{{Article history

| dykdate = 9 December 2013

| dykentry = ... that '''[[Convolutional neural network|convolutional neural networks]]''' have achieved performance double that of humans on some image recognition problems?

}}

{{WikiProject banner shell|class=C|

{{WikiProject Cognitive science}}

{{WikiProject Computing|class=C|importance=low|needs-image=yesMid}}

{{WikiProject Artificial Intelligence|importance=Mid}}

}}

{{User:MiszaBot/config

| algo = old(90d)

| archive = Talk:Convolutional neural network/Archive %(counter)d

| counter = 1

| maxarchivesize = 125K

| archiveheader = {{Automatic archive navigator}}

| minthreadstoarchive = 1

| minthreadsleft = 5

}}

== Inaccurate information about Convolutional layers ==

== Feature Maps ==

Need to introduce what feature maps are for nontechnical readers. <!-- Template:Unsigned --><small class="autosigned">—&nbsp;Preceding [[Wikipedia:Signatures|unsigned]] comment added by [[User:Shsh16|Shsh16]] ([[User talk:Shsh16#top|talk]] • [[Special:Contributions/Shsh16|contribs]]) 18:24, 15 February 2017 (UTC)</small> <!--Autosigned by SineBot-->

== Non-linear Pooling ==

Convolutional layers do not do convolutions. They do what is called "Cross correlation" in DSP, which is different than the statistics definition of cross correlation.

It says in the article: "Another important concept of CNNs is pooling, which is a form of '''non-linear''' down-sampling."

https://en.wikipedia.org/wiki/Cross-correlation

This article says multiple times that the convolution operation is being done, and it links to the convolution article

I don't think this is correct. There are pooling techniques, like average pooling which is mentioned in this same section, which are forms of linear down-sampling. I would remove the "non-linear." [[Special:Contributions/194.117.26.63|194.117.26.63]] ([[User talk:194.117.26.63|talk]]) 15:06, 13 May 2016 (UTC)

https://en.wikipedia.org/wiki/Convolution

This is misleading because it does not do this operation linked in the article. It does the operation linked in the cross correlation articles.

== Plagiarism in "Layer patterns" ==

-AS

== Inacurate information: Convolutional models are not regularized versions of fully connected neural networks ==

The text seems is copied from https://cs231n.github.io/convolutional-networks/#layerpat without any attribution <small><span class="autosigned">—&nbsp;Preceding [[Wikipedia:Signatures|unsigned]] comment added by [[User:Jkoab|Jkoab]] ([[User talk:Jkoab|talk]] • [[Special:Contributions/Jkoab|contribs]]) 01:41, 8 June 2016 (UTC)</span></small><!-- Template:Unsigned --> <!--Autosigned by SineBot-->

In the second paragraph of the introduction, it is mentioned that "CNNs are regularized versions of multilayer perceptions." I think the idea is inaccurate. The entire paragraph describe convolutional models as regularized versions of fully connected models, and I don't think that is a good description. I think the idea of [[Inductive bias | inductive bias]] would be better then that of regularization to explain convolutions.

:Indeed. Deleted copyvio text, see below. [[User:Maproom|Maproom]] ([[User talk:Maproom|talk]]) 09:55, 8 June 2016 (UTC)

== Copyright problem removed ==

I would also suggest merging the section "Definition" into the introduction. The definition section is only two sentences and it feels it would be better placed at the introduction.

Prior content in this {{#ifeq:{{NAMESPACENUMBER}}|119|draft|article}} duplicated one or more previously published sources. The material was copied from: https://cs231n.github.io/convolutional-networks/#layerpat. Copied or closely paraphrased material has been rewritten or removed and must not be restored, ''unless'' it is duly released under a compatible license. (For more information, please see [[Wikipedia:COPYRIGHT#Using_copyrighted_work_from_others|"using copyrighted works from others"]] if you are not the copyright holder of this material, or [[Wikipedia:Donating copyrighted materials|"donating copyrighted materials"]] if you are.)

== Introduction ==

For [[Wikipedia:Copyrights|legal reasons]], we cannot accept [[Wikipedia:Copyrights|copyrighted]] text or images borrowed from other web sites or published material; such additions will be deleted. Contributors may use copyrighted publications as a source of ''information'', and, if allowed under [[fair use]], may copy sentences and phrases, provided they are included in quotation marks and [[WP:CS|referenced]] properly. The material may also be rewritten, providing it does not infringe on the copyright of the original ''or'' [[Wikipedia:Plagiarism|plagiarize]] from that source. Therefore, such paraphrased portions must provide their source. Please see our [[Wikipedia:NFC#Text|guideline on non-free text]] for how to properly implement limited quotations of copyrighted text. Wikipedia takes copyright violations '''very seriously''', and persistent violators '''will''' be [[Wikipedia:Blocking policy|blocked]] from editing. While we appreciate contributions, we must require all contributors to understand and comply with these policies. Thank you. <!-- Template:Cclean --> [[User:Maproom|Maproom]] ([[User talk:Maproom|talk]]) 09:55, 8 June 2016 (UTC)

"only 25 neurons are required to process 5x5-sized tiles". Shouldn't that be "weights" and not "neurons"? Earlier it said "10,000 weights would be required for processing an image sized 100 × 100 pixels". [[User:Ulatekh|Ulatekh]] ([[User talk:Ulatekh|talk]]) 15:53, 19 March 2024 (UTC)

== Suggestion: Move the section "Regularization methods" to a new page ==

:Absolutely, you're right. I was going to ask the same question. 25 weights for each neuron in the second layer from each neuron in the input layer, and all these 25 weights don't vary as the filter is slid across the input. Do you want to make the correction or should I, since the original editor is not responding? [[User:Iuvalclejan|Iuvalclejan]] ([[User talk:Iuvalclejan|talk]]) 22:47, 25 January 2025 (UTC)

The methods listed here are applicable to deep learning in general.

::I made the change. [[Special:Contributions/2600:6C5D:577F:F44E:B9B2:E830:3647:8315|2600:6C5D:577F:F44E:B9B2:E830:3647:8315]] ([[User talk:2600:6C5D:577F:F44E:B9B2:E830:3647:8315|talk]]) 14:20, 27 January 2025 (UTC)

This topic should be moved into a new page. [[User:OhadRubin|OhadRubin]] ([[User talk:OhadRubin|talk]]) 06:38, 27 November 2018 (UTC)

== Big picture ==

== Parameter Sharing Clarifications ==

Why are convolutional NNs (or networks with several Convolutional layers as opposed to none) more useful especially for images, than networks with only fully connected layers? You mention something about translational equivariance in artificial NNs and in the visual cortex in brains, but this is a property of the neural network, not of its inputs. It's a way to reduce the number of weights per layer, but why isn't it universally useful (for all inputs and all output tasks), and why is it better for images than other ways of reducing the number of weights per layer? [[User:Iuvalclejan|Iuvalclejan]] ([[User talk:Iuvalclejan|talk]]) 23:50, 25 January 2025 (UTC)

In the "Parameter sharing" section, "relax the parameter sharing scheme" is written, but what this actually means is unclear. <!-- Template:Unsigned --><small class="autosigned">—&nbsp;Preceding [[Wikipedia:Signatures|unsigned]] comment added by [[User:Ephsc|Ephsc]] ([[User talk:Ephsc#top|talk]] • [[Special:Contributions/Ephsc|contribs]]) 16:22, 27 September 2019 (UTC)</small> <!--Autosigned by SineBot-->

==Wiki Education assignment: Linguistics in the Digital Age==

== What is convolutional about a convolutional neural network? ==

{{dashboard.wikiedu.org assignment | course = Wikipedia:Wiki_Ed/University_of_Arizona/Linguistics_in_the_Digital_Age_(Spring_2025) | assignments = [[User:AshlaMaOmao|AshlaMaOmao]] | start_date = 2025-01-15 | end_date = 2025-05-09 }}

<span class="wikied-assignment" style="font-size:85%;">— Assignment last updated by [[User:FblthpTheLost|FblthpTheLost]] ([[User talk:FblthpTheLost|talk]]) 00:10, 8 May 2025 (UTC)</span>

The article fails to explain what the connection between CNNs and [[convolution | convolutions]] are in any meaningful way. In particular, convolutions don't act on vectors; they act on functions. Comparing with the equation on the page for convolutions, there's obviously something analogous. --[[User:Stellaathena|Stellaathena]] ([[User talk:Stellaathena|talk]]) 16:51, 14 December 2020 (UTC)