Image Classification Using Swin Transformer

[@Sunil-ez1hx]

Impressive content & amazing way of explaining

[@CodeWithAarohi]

Thanks a ton

[@pifordtechnologiespvtltd5698]

Amazing work done. Keep it up

[@CodeWithAarohi]

Thanks a lot 😊

[@soravsingla6574]

Hello Ma’am Your AI and Data Science content is consistently impressive! Thanks for making complex concepts so accessible. Keep up the great work! 🚀 #ArtificialIntelligence #DataScience #ImpressiveContent 👏👍

[@CodeWithAarohi]

My pleasure 😊

[@ABHISHEKCHINCHOLE-im8gn]

mam u have uploaded the img of daisy flower and in o/p it shows dandelion flower🌸 🌸

[@palurikrishnaveni8344]

Your videos are very helpful for my research work. Thank you so much madam

[@CodeWithAarohi]

It's my pleasure

[@palurikrishnaveni8344]

I will implement this too with my datasets Thank you madam

[@zappist751]

Thank you! Can you please show us how to use the pre-trained swin v2 b weights for this?

[@CodeWithAarohi]

Will cover soon

[@zappist751]

@@CodeWithAarohi Thanks madam

[@nandiniloku7747]

IT WILL BE VERY MUCH USEFULL @@CodeWithAarohi

[@Musaddique_MMALI]

@@CodeWithAarohi Have you covered it?

[@fayezalhussein7115]

You are amazing, thanks alot Aarohi

[@CodeWithAarohi]

Welcome

[@zahidkhan-ee4th]

Thank you for sharing this information; I need to know how to use CNN with Transformer.

[@Sunil-ez1hx]

Thank you soo much mam for this amazing video

[@CodeWithAarohi]

Thanks for liking

[@Sana_Premchand]

Detailed explaining. I loved it. Can you please make a video on classification problem based on 'Video dataset 'using Swin transformer?

[@CodeWithAarohi]

Noted!

[@soravsingla6574]

Very well explained

[@CodeWithAarohi]

Thanks for liking

[@НиколайНовичков-е1э]

Thank you, Aarohi!

[@CodeWithAarohi]

Welcome 🙂

[@RahulChatterjee-lu7mv]

great video ,can you tell the way of doing the same on a dataset ,where there will be a collection of images

[@陈子昂-v4t]

maybe you can check the vedio about vit and you will find the web of dataset

[@居陳]

Nice video! Can u explain how Deit works, and make a coding demo video.

[@CodeWithAarohi]

Yes, soon

[@umamahesvari1309]

Awesome video mam. Can you make a video for image segmentation using swin transformer?

[@CodeWithAarohi]

Sure, Soon.

[@sahaj2805]

Amazing explanation..Can you also please make a video of using hybrid architecture i.e. CNN with pretrained swim transformers to solve the image classification problem , thanks in advance:) Also videos explaining in detail the vision transformers and swin transformers research papers.

[@CodeWithAarohi]

Will try

[@somabanerjee5027]

Can you please share the detail of your coding platform. Hardware and operating system.

[@soravsingla6574]

Code with Aarohi is Best KZbin channel for Artificial Intelligence #BestChannel #KZbinChannel #ArtificialIntelligence #CodeWithAarohi #DataScience #Engineering #MachineLearning #DataAnalysis #BestLearning #LearnDataScience #DataScienceCourse #AytificialIntelligenceCourse #Codewithaarohi #CodeWithAarohi

[@growwithfuyad4497]

how many total numbers of prams we using in this model?

[@nandiniloku7747]

mam can you please show how we can use pre trained weights to swin transformers

[@SS-zq5sc]

Hello, another great video. Thank you very much. If possible, could you also cover image segmentation using swin transformers. Using a custom dataset.

[@CodeWithAarohi]

Yes, soon

[@sahaj2805]

In Swin transformers also we are resizing the image to 224x224 like in vision transformers , but you said that Swin transformers can handle large size images , can we pass directly the input image in Swin transformer of resolution type 1024x768 ?

[@CodeWithAarohi]

Yes, in SWIN Transformers, unlike traditional vision transformers where the entire image is typically resized to a fixed size like 224x224, you can indeed directly pass images of larger resolutions such as 1024x768. SWIN Transformers are designed to handle such large images efficiently by breaking them down into smaller patches and processing them hierarchically.

[@murphy1162]

great job !!! would you mind making a vedio that using Swin Transformer for object detection ?

[@CodeWithAarohi]

Sure, I will try to make it soon.

[@Electro_words_with_sandhya]

well explanation! I tried this for large data, but it does not work well in terms accuracy and loss. Any solution

[@CodeWithAarohi]

As you mentioned the large dataset. Are you sure that the quality of dataset is good because the quality matters more than quantity.

[@Electro_words_with_sandhya]

@@CodeWithAarohi thank you, but I have been using this for face forensics. It does not get trained well, it just gets time our. Now the dataset is 70K images, and training is not even starting. Will it not be working in Google Colab.

[@kartikpandit7438]

Can make a video on how to use Swin transformer v1 or V2 for object detection with bounding boxes and annotations

[@CodeWithAarohi]

Sure

[@NandanChhabra91]

Another great video Aarohi. Thank you! Question : Can you share any references to follow where I need to classify a video dataset using Swin Transformer ?

[@marzi869]

keep going, ok.

[@CodeWithAarohi]

Sure, Thanks

[@abubakarsaleem5167]

Is it compulsory to load the image size of 224 for swim transformer network?

[@CodeWithAarohi]

No, it's not compulsory to load images of size 224 for SWIN Transformer network. The SWIN Transformer architecture is flexible and can handle various input sizes.

[@mohdUbaidWani]

mam plx can u make a separate video how to run a githu repositry in anaconda prompt...please🙂

[@CodeWithAarohi]

Sure, Will do it.

[@fayezalhussein7115]

dear would you explain transfer learning for swin

[@CodeWithAarohi]

Will try

[@nadeemchaudhary16]

Mam, please provide the code for pre pre-trained swin transformer.

[@CodeWithAarohi]

I haven't prepared that yet.

[@suvarnabhat7212]

yes ma'am plz provide code for pre-trained swin transformer

[@DeepakKandel-go3ff]

Can you do it for Image Segmentation?

[@CodeWithAarohi]

Will try to cover.

[@shatadruroychowdhury6319]

Why is the embedding size 96?

[@CodeWithAarohi]

The choice of an embedding size of 96 in the SWiN Transformer model is a design decision that balances computational efficiency and model performance.

[@abdelrahimkoura1461]

Hi Aarohi, still it would kindly if you allow me to use your data. You can put your data in the repo