Adding the EFDM method, CVPR2022

[YBZh]

Thanks for your excellent codebase of Dassl.

Based on Dassl, I have validated a novel method, named EFDM, which has been accepted by the CVPR2022 （the paper can be found at: https://arxiv.org/abs/2203.07740）

In this pull request, I am adding the EFDM into Dassl to make it more powerful. Thanks for your time and consideration.

[KaiyangZhou]

Congrats! And glad to know that Dassl is helpful to your research! :)

I've checked the submitted files and they look fine.

So before merging the PR, I'd like you to run linter.sh to format the code and make sure there is no lingering warnings (let me know if you have done this).

Feel free to create a new issue titled "how to run EFDM (CVPR'22)" and provide instructions. Practitioners who are interested in using your work would want to know.

[YBZh]

Thanks for your quick reply.

I have formatted the codes with the "linter.sh".

I will create a new issue titled "how to run EFDM (CVPR'22)" and provide a link to the repository: https://github.com/YBZh/EFDM, where detailed guidelines will be available.

Thanks again for your time.