We need to cover knowledge transfer of this model to be reused in other networks. like using this network as a backbone to another network. ex: image segmentation using swintfv2 pretrained weights as encoder and do the decoding using UNET decoder

thanks for explaination. I want to combine the SWIN with YOLOv8 any help on this topic.

what happened to b in P2.3

Thanks for the excellent videos. I have a question how can I train my custom data using that swin transformers

Amazing video. Thanks a looooooot:)

Thank you

I have done a bunch of deep learning studying on youtube. This has to be the course with the most value I have ever seen. This course is amazing! Thank you so much for your contributions!

A question about the code: self.tau should be a tensor of size: "self.tau.shape = d". where d is the number of heads.

The GitHub repo that you provided has PatchMerging not the PatchMerging_conv. Can you tell the difference between both and the need to use different methods?

How can a Swin Transformer backbone output multi-level feature maps like ResNet?

great explaination

Very well explained. Thank you for making such a good tutorial 👍

Can this be used on character images of 32*32*1 dataset

thank u sir made it easy well explained can i use the final code as backbone to extract the initial features from the image and pass it to another transformer ?

hey can I use it to build my own ocr model?

I need some help related to object detection using swin Transformer.Can you please share your email address.

please send your email address. want to communicate related to object detection using swin transformer.

This series was an excellent explanation of the Swin transformer; thank you for all your effort, What parameter or Defs should I change if the input is 3D images?

Hats off to your effort to explain this SWIN beast

Dear i need your help in swin tansformer project, how i can contact with you ?

Thanks so much!!

Again, thanks so much, with your teaching it is impossible to NOT get it ;)

Very nicely explained. Thanks so much!

Thank you so much for creating this Swin Transformer tutorial series! Your clear explanations and well-structured content have been incredibly helpful in understanding the intricacies of Swin Transformers. Your effort is greatly appreciated, and I'm sure many others like me have benefited from your expertise. Keep up the fantastic work!

Thank you for your perfect teaching in ditals

Good Video, FYI I don't think the line to plot the predicted values is shown. For others reference, these two lines should allow the plots to work. plt.plot(X.numpy(), Y.numpy(), 'r+', label = 'Y') # This is shown in the video plt.plot(X.numpy(), Yhat.detach().numpy(), label = 'Yhat') # This needs to be added to plot the prediction line

I really wish you would've taken this course further

Fablous series of videos. Jsut watched everything. Thank you for such valuable content!

can you please explain how to add swin transformer in the backbone of the YOLO v5 model

Thank you very much. You are the best

what about the segmenatation?

Very good, clear explanation, thanks for making this!

How to trained the swin transformer for the rectangle image i.e 64x512x3

your explanations are so clean and detailed. thanks very much

Nicely Explained..! Please bring some latest SOTA and Benchmark models for Image Segmentation also.

This is the best video of mask in swin I've ever seen! Salute and hope to see more other useful videos

Great Video, Thank you so much! Can you please make a video for 3D SwinUNETR?

I appreciate the lesson. I wonder how you can get the formula for nonlinear regression because in the lesson I see some figures but the way to get formula from figure you do not mention.

For self.upper_lower_mask and self.left_right_mask, can I use register_buffer instead of nn.Parameter and setting requires_grad=False?

Hello AI OpenCourseWare, First of all, let me thank you for making such a wonderful playlist that covers each and everything in a very good and easy way. Secondly, I was struggling with the swin transformer and which was a big block in my research work. Now that I have understood many minute details and concepts from your videos, I am kind of comfortable now. I request you to please come up with more videos on computer vision topics. Also, please bring the latest research work implementation like Hypercorrelation squeeze network, 4D convolution, Swin4D, cost aggregation with transformer, and so on... Once again thank you for helping us out in understanding the code and minute concepts. Thank you..!

This is amazing hard work thank you so much.

Thanks for providing these videos🙏👌

What an amazing explanation👌👌, thanks a lot🙏🙏

what will be the value of C if input image is 256x256, what all must i change for it work with 256x256x3 images

Just woww what a wonderful explanation loved it <3 you earned a sub mate!

thanks!!

Appreciate your efforts! Thanks :)

Amazing explaination. Thank you!

you made swin transformer easy for us, thank you so much

Best Explanation I've Ever Seen. Thank you so much sir.