Pooling Layer in CNN | MaxPooling in Convolutional Neural Network

[@Abhishek-pb8kt]

Can keep listening to the content for hours without stress or anxiety :D and no drop in enthusiasm or learning. Covering every important detail methodologically.

[@swiss8943]

I wish I could listen to it for hours, but I have my exams right after 2 hours

[@wanderlust_womaniya]

Thank you so much for clarifying this in such a simple language and specially with showing websites that give visual demonstration of such concepts! 🙏🙏🙏🙏

[@pravinshende.DataScientist]

I have learned a lot much from this channel .. Thank Nitish sir!

[@rishabhvyas856]

One of the best teacher i ever found ❤ I feel so lucky that i found you. Thank you so much sir i really appreciate you🙏

[@AryanSingh-fe5uy]

Sir your teaching flow and method is awesome

[@balrajprajesh6473]

First I like, then I watch!

[@meerzamanatali8914]

Same 😊

[@Vid_Lumina]

Us moment

[@yashgupta7604]

Sir you are one of the best teachers i have ever studied from .Keep shinning sir. I am very lucky that i found you.Keep showering knowledge sir 😇

[@chetanmundhe7899]

Am 3rd, I just started watching but thought of putting comment first.... You are greattt broo... Lots of blessings

[@pravinshende.DataScientist]

every time you properly understand the why concepts that why we get the concptual clarity .. so ratta marne ki jarurat nahi hoti

[@sahuchiragshyamlal3684]

problems with convolution layer: i) memory issue ii) translation variance => features are location dependent solution: i) strides solve first prob only ii) pooling solve both prob relu apply on feature map so non linear feat map then apply pooling(downsampe feature map) model1.add(MaxPooling2D(pool_size=(2,2),strides=2, padding='valid')) maxpooling in receptive field(pool_size) keep dominance feature and discard low level details 15:00 min for dimensions No training require during backpropagation Advantages: i) reduce size of feat map ii) Translation invariance iii) No need of training bcz its just aggregate opetaion iv) enhance feat only in max pooling 24:00 global max and avg pooling Disadvantages: i) Translation invariance not used in Image segmentation ii) Loose lot of info

[@debojitmandal8670]

But from wht i understood is that pooling will reduce your image size therefore your loosing some information how do you solve that

[@samaysingh5794]

@@debojitmandal8670 you are right but it is like generalizing i.e. getting overall idea it is some what removing overfitting which we can call translation variance in CNN

[@algocoholic2494]

2nd time.... 1st view and 1st comment 🤣🤣🤣🤣🤣...love you sir.... Hope you have read the form that I filled

[@piyushpathak7311]

Sir please complete this series as soon as possible..

[@gauravyadav3315]

LEGEND 🔥

[@savyasachi6988]

very nice pooling explanation

[@AlAmin-xy5ff]

You are really awesome sir!!!!!!

[@guru_bro]

Very good 👍

[@homeboy1733]

Awesome Sir

[@rb4754]

First I like then I watch the video and then I comment...

[@Pawan-pm4tw]

Salute sir❤️

[@gurpreetkaur-pf1bf]

Thku from heart ❤️

[@farhadkhan3893]

you discussed a problem in the previous video of Padding and strides that some information is lost while featuring the images so you applied the padding to maintain the size of the image. so my question is, isn't it a problem that some information is lost in pooling?

[@narendraparmar1631]

help a lot Thanks

[@mr.deep.]

Best

[@mihirthakkare504]

Santre colour ka billa❤ in video is lub

[@tanveerbashir8393]

Thank you🥰😍😇

[@yashjain6372]

awesome video. Doubt : 20 Min But why we are not getting shape of 8 by pooling ?

[@sandipansarkar9211]

finished watching

[@AmitKumar-kt8oc]

Amazing 😍

[@vairab1876]

What is the size of kernel in conv2d_1 ? there are 32 kernel , is it (3,3) or (3,3,32) ?

[@vivekacharya3652]

Main first😄😄

[@elonmusk4267]

great

[@rahuljha3686]

✓ done

[@rachitsingh4913]

Can we use padding and pooling together ? Because by using padding we are actually stopping our features from loosing information whereas by using pooling we are loosing some information . Can someone please clear my doubt ?

[@chetanmundhe7899]

Quick question... Do you work for any company or work for yourself?

[@nipunsingh7740]

If possible, do upload your notes too along with the videos

[@shashankshekharsingh9336]

legend🔥🔥

[@ppal6329]

The best

[@varunnnwtf]

JODDD❤❤

[@muhammadyaseen2856]

Sir please create playlist on -- Computer Vision and Image Processing --

[@charanpoojary4804]

great

[@ebrahimhaquebhatti3258]

2:50 Isn't it 224x224x3?

[@salihamirza456]

BEHTREEN!!

[@divyakarlapudi]

best

[@harshumokal5736]

Please do something on object detection

[@divakarsaraswat7820]

Sir, Data analyst k liye full road map machine learning ka

[@pkumar0212]

👌

[@MN-gn7lt]

Sir course jaldi launch karo na

[@rohitpotluri4378]

CFBR